Home

PDF Document

Contents

1. 5 Design space 51 5 2 Desienspate d sca ia duh ARE ace DI OVERVIEW sa ca DS R e 5 112 Design rationale 5 1 3 Notationsandrepresentation 5 14 Analysis methodology and guidelines A comprehensive look at the navigation methods in digital media 521 Digital document Current design overview Future design suggestions 5 22 Picture Interactivemap Current design overview Future design suggestions 23 23 29 30 34 37 39 43 43 43 44 50 53 60 61 61 61 66 Contents 5 23 Digital audio Current design overview Future design suggestions 5 24 Digital video Current design overview Future design suggestions 6 Evaluation block diagram 6 1 Structure 62 Notationsand representation 6 3 Generation from design Space 6 4 Evaluation methodology 7 Validation 71 Task amp Setup 72 Participants 7 3 User test program 74 Results 75 Analysisandconclusions 7 51 Digitaldocument 7 5 2 Interactive map 75 3 Digital video 8 Summary and future work 71 71 71 75 75 75 79 80 82 84 85 89 89 95 96 107 108 108 112 114 117 vii viii Contents 81 Summary and contributions 117
2. juoumoop eji3rp 107 Sur 015s paseq aje1 IIYM esnoui JO WEIZEIP ADO UONHEN LAJ 9 Y JINBL Mi af Bok wag 27759 de Kr y Ka er ez a xo TOMADO geng ein M mern Pa d Pur ON Mz gt LCD j A Evaluation block diagrams for digital document 128 soua nr a Tess pan y wurine pe men A AN mnra wenmod N ES j Sa MN mein a Dm m d 9 T d x nan y f urn W an e uopurg N 1 sag ien wd y Ps mad be q a TEG L mme N L ems A NT j ir A gt j Na di WPG li jueumnoop TEYSTp 10 gum fods uognq umop dn preoq43y Jo ure18erp porq uorenfeag Z Y 31314 129 jueurmnoop ejiBrp 103 spreuqumy shay more preoq e jo ure18erp poq uonen eAq 8 Y onSrq A Evaluation block diagrams for digital document 130 juounoop ejiBrp 103 uo3nq umop aged dn a8ed preoghay jo ure18erp od uonenpeAq 6 y 31313 l wen e ma Y ug Ne nanye j gen N en 7 n Thomos y met y L wont RE gt aki TE N s d d aa d 7 Lang ka wag NV a ae perq CM e e ER muomba ar Bo peu Ae e 131 juounoop eit2rp 10 NAML JO ure18erp od uonen eAq 01 Y 31314 A Evaluation block diagrams for digital document 132 jueurmnoop ejiBrp 103 uonoung puy xoq Yoreas jo ure12erp porq UOneEN LAJ EV 31314 m wopurg Y f a A dung a mue we N Z Je a Sun 0 7 wtu Y
3. Access ege machann promos Oe wars daf irem ure HE anve Oper even Me Ek ano Tears parda m fe rents of fe aser e Support any type of dete In not meno war s m nr md qr rt nach mm e Werbstitiao Undepembeme o non ran ah ete mor beur independent on Wine Als age Lrmien are requis mc meni TT ami monrag U amar Figure 7 2 Test program screenshot search box in digital document 100 7 Validation Task 1 Find the graaf of Theorem 5 17 for this tank you can only une siider to navigate Press FI te laad the document and start time recording press F2 after you Pad he target Tate completed pwose presi Me Cortieus button Renn Con Dra 2900 Aquin if wa omsider the Bra VASK tio ia 10 0 1 00 0 13 C71 0 ETE NET TEL Then we muit do the anme work for the tibren ether alternatives ere sponding to cycles feum a r mad di This eech to the amp llreing wazi Se parte for a c and d tan mm tn tn la tan tan ta tm ton lee me m len 7 014 mm 019 0 em or hy lay deam om mde dan pe sea bay my tan mf The express he 2 difkemani frum the uiers bwason nonempty cycies st b mast po s tel ner Jean tbe large tyde while other arran cycle tas am through the amail cycle onby The resulting rent reet pan f Y deecribing weights of en correspon to the inim of these rem tege parts namely Nan ma tm a en mm A taa L saj ll Kom m tant m tan nm san antan ml For the bra VAS
4. 8 2 Future WORKS w eot Co epa o eod 119 A Evaluation block diagrams for digital document 121 B Evaluation block diagrams for picture interactive map 135 C Evaluation block diagrams for digital audio 145 D Evaluation block diagrams for digital video 151 E Raw test data 161 Bibliography 171 Index 175 List of Figures 3 1 3 2 3 3 3 4 3 5 3 6 3 7 3 8 3 9 3 10 4 1 4 2 4 3 Smartplayer Elastic audio slider The dimensions and the taxonomy in design Space f aa RTE da e LIN R en Summary of navigation methods for digital document Gay o a ee es Summary of navigation methods for pic ture interactivemap ix 4 4 4 5 5 1 5 2 5 3 5 4 5 5 5 6 5 7 5 8 5 9 6 1 6 2 6 3 7 1 7 2 7 3 List of Figures Summary of navigation methods for digital aco A VA os ER xb IE dea S 38 Summary of navigation methods for digital Video s tin Sa eel Cie EE 41 Design space overview 45 Notation of design space for digital document 54 Notation of design space for picture inter active map un een Ae 55 Notation of design space for digital audio 56 Notation of design space for digital video 57 Design space for digital document 64 Design space for picture interactive map 69 Design space for digital audio 73 Design space for digital video 77 Evaluation block dia
5. Intelligible audio feedback The authors developed a method to provide audio feedback during timeline slider scrolling A small snippet of audio 1 2 seconds starting at the current slider position will be played in full before the play of the next snippet no matter what the user does in between In addition the rate at which the snippets are played is determined by the scrolling speed 19 20 User study of rate position control scrolling User study of rate position based input devices User study of scrollbars tap and drag and touch n go 3 Related work 3 3 User study and evaluation Some of the related work described above included user tests to evaluate the proposed designs and to compare with the existing ones Besides there is some other previous work which did user studies to evaluate different naviga tion methods The results will be presented in the follow ing part and be used to support the proposed design space in the later parts Hinckley et al 2002 did an experiment to compare the performance of IBM ScrollPoint rate control scrolling IntelliMouse Scrolling Wheel position control scrolling and two enhanced scrolling wheels with acceleration al gorithms one at 3 lines notch and one at 1 line notch integrated The results presented a crossover effect be tween devices depending on the scrolling distance At short distances the rate control scrolling was significantly slower than the wheel techn
6. fi M mopurg Mo T A 3 arm 4 d puras LIR Haw A L anz au w Fi En TOS S IER Ga min y LI Bleu d e ene NE T os N GE i l eie mapa y MA d VERE og e ge roun prrurrkss avwe i anne 1 penny j A an pnl f LOS 312126 Tonie partt gL 139 deur sAnIETAJUI a1njord 103 TEULIA Uuaar2s yono SNJAIS asnow Jo ure18erp 190 q uonenjeAg g mS an N B Evaluation block diagrams for picture interactive map 140 deu aan oe19jur AMIAA 107 mo UT WOOZ 10j Adi pue SUTAOUI 107 S a moure preoq4ay jo wer3erp 32010 uonen eAq eg am ann od 1 who DEDE mepury ranas L ng twatrn Aa Lem 141 deur aAnoerojur aqnjord 103 xons of Jo ure18erp xpo q uonen eAsq 9 g am B Evaluation block diagrams for picture interactive map 142 dew aAnoerayur AMPIAA 101 NAM JO ure18erp 32010 uonenjeAq 7g enSrq aa N ope OC j F A Gre v Y amog 4 0 1 linad le La LAN eonim E j ko y E Me df a Be D Dau el He mde r p EECH s pr y d M ya ee a E 4 ng aa Pn Y weie yon so wopurg ass ka Una od Y gt SI ELEM e moauanbis on mn passe ET 24 TREES a HI uogem sts peeeepest 143 dew aAnoeTajur ampid 101 xoq VoTeas jo ure12erp po q uonen eAg g g exnSrq CN STILE e Appendix C Evaluation
7. screen objects when navigating large space contents such as maps Figure 3 6 Halo accomplishes this by surrounding off screen objects with rings which reach into the border re gion of the display window Users can infer the off screen location of an object by estimating the center of the ring The user study showed that users completed tasks 16 3376 faster using Halo off screen indicators than using Ar row off screen indicators Figure 3 6 Halo Karrer et al 2008 developed DRAGON DRAGable Object Navigation a direct manipulation interaction technique for the frame accurate in scene video navigation tasks Unlike the traditional timeline slider navigation which has a non linear mapping from movie time to object position that leads to indirect unnatural and inconsis tent navigation DRAGON enables user to quickly and precisely drag the interested object in scene to a specific desired position As shown in Figure 3 7 an object e g a Halo Dragon 15 16 3 Related work car a passenger or even a piece of paper on the ground can be dragged along its movement trajectory The eval uation showed that DRAGON significantly reduces the task completion time for in scene navigation tasks by an average of 19 42 compared to a standard timeline slider And qualitative feedback showed that users felt more natural when using DRAGON to navigate the videos Be cause of the frame locating accuracy of direct manipulation in
8. which proves the hypothesis Therefore it is concluded that for digital video discrete random access has advan tage against continuous sequential access Chapter 8 Summary and future work This chapter provides a retrospection of what have been done in the work summarizes the problems addressed in the work and contributions for the field of human computer interaction After that some future work idea is proposed for going deeper into the theoretical research of navigation and design space 81 Summary and contributions Overall speaking the work is a theoretical analysis of the navigation in digital media in the form of creating a de sign space By introducing the background of navigation in digital media and design space in 2 Background and by reviewing the related work of navigation design analy sis and design space in 3 Related work we know that there had been no previous work to create a design space for the navigation in digital media and to have a compre hensive look at the navigation methods used in different types of digital media up to now Thus this work system ized the analysis of navigation in digital media by creat ing a design space to describe the navigation in an abstract way which classifies navigation methods according to the navigation properties Furthermore the work provided a 117 Summary of the work 118 8 Summary and future work series of analysis methodologies to systematical
9. zapa d nou 123 jueurnoop ejiBrp 107 req oros pe ueque reuquinuj usams yono SNIAIS asnouw jo ure38erp porq uonen eAq 7 y am ma P e keet Aal T wf ge X nm K Nei J nd t Ken monn A nan A ee A Barrad mg Ben HS di geg Vie e m e Fa EN PE Gate SC L d se L Pte hM Be rai u pubes A Evaluation block diagrams for digital document 124 jueurnoop eji3rp 10j spreuquny USIMS qonoj SNIAIS asnoui jo ure18erp porq uonen eAq CN am ge wopmg am sg mambo L eene N LL _ ee M a M 9085 r na wu bu ti pat 2 amv y Dora P A r as nan A d wurm ueg mopang a manm H pout 125 juoumoop eni3rp 103 Sexp pue dej Uaar2s yono Jeng asnow jo ure18erp porq uornenfeAq p y amSuy A Evaluation block diagrams for digital document 126 juoumoop ey 2g1p 107 Sui 015s aan YAN Terp Aol paseq uonisod JaayM asnouw jo ureIderp xpo q uonen eAq Ev 310311 M ww Z mops P T p a Py furscg nente MES ponm A A 2 M ge ka E Pa i tanun SEN r n gaufres d j m E Se mem e e f w g YO ff wnso d N i j NE N P j no f d TN m j AST pas I d Xi ka 4 e ine gt a wf nan v gen QUU PES E m uopurg N wg L neng mr y a Bia 7 E 3 PTA gt d rn i mp5 ner j A Pow t ba gt BG PRD 127
10. Keyboard up down button Scrolling The levels are classified by the preferences The naviga tion methods in the different levels have different naviga tion properties while those in the same level have the same properties according to the evaluation criteria of this work Therefore the navigation methods in the same level are be lieved not to have significant performance differences al though there may exist some minor differences An example evaluation task 87 6 Evaluation block diagram ure18verp Yyoo q prepuejs o durexo uy 9 an By Chapter 7 Validation A user test is done in this thesis work The purpose of the user test is to get empirical data to verify the correctness of the hypothesis for designing the layout of the design space the preference proposed when using the evaluation block diagram to measure the navigation methods and thus to verify the correctness of the results derived from the evalu ation block diagram analysis However because of the time and resource limitation the user test done in the work can only verify parts of the framework which are regarded as the most important aspects More validation work will be done in the future work 7 1 Task amp Setup The user test is a quantitative experiment which is run by a program to measure the task completion time The test ad dresses three types of digital media which are digital doc ument interactive map and digital video For each of the
11. Stylus Touch screen Fast forward Rewind button frame skipping Relative input Rate based control Tool based navigation Discrete amp Sequential access targeted playback position search not possible 6 Shuttle wheel Spring loaded Relative input Rate based control Tool based navigation Continuous amp Sequential access 39 40 4 Navigation of digital media 7 Jog dial Click wheel Relative input Position based control Tool based naviga tion Discrete amp Sequential access 8 Dragon Absolute input Position based control Direct manipula tion Continuous amp Sequential access targeted playback position search not possible 41 4 2 Navigation methods OSPIA e ISIP 107 spoujeur uoneBTAeU Jo Kreurumng et JINBIJ Chapter 5 Design space With the taxonomy of navigation given the proposed de sign space will be presented in this chapter An overview of the design space will be taken at first with the notation and representation to be explained Some analysis methodolo gies and guidelines about how to use the design space are going to be discussed afterward After that a comprehen sive look at the navigation methods in the four addressed types of digital media will be taken By using the design space the current navigation methods and some suggested future designs will be analyzed 5 1 Design space 5 1 1 Overview Based on the taxonomy given in the previous chapter a table f
12. nutzertest der die Richtigkeit des Designs der Grundstruktur und der Ergebnisse des Evaluierungstools berpr ft Er gibt die Gewissheit dass das Design und das Evaluationsstruktur sinnvoll f r die Entwickler ist xxiii Acknowledgements This thesis would have never been accomplished without the contribution of many people First of all I would like to thank Prof Jan Borchers for offering such good and interesting courses which lead me to the HCI field Many thanks to Thorsten Karrer and Gero Herkenrath for offering me this interesting and challenging thesis topic Throughout the thesis work You gave me a lot of idea guided me in many aspects from theory user test design to coding techniques I learned so many things from you in this work Thanks my advisors Thanks to my friends who are alway caring about the progress of my thesis work talking about the topics with me and gave me help when I have difficulties Because of you my stress from the work is released to a large extent I would also appreciate my thanks to all the friends and students participated in my user test Thanks for your time to do the test in which you experienced some tough tasks Special thanks to my family Thanks for your support which makes me possible to finish the study in Germany And thanks for your understanding of my lack of attention due to the stress of writing this thesis XXV Conventions Throughout this thesis we use the following
13. three media types several selected navigation methods are tested Each user is asked to do all the tests for the three media types 89 90 Task 8 Setup for digital document 7 Validation Part 1 Digital document Task Find a required paragraph figure theorem or defi nition etc in a pdf document Selected navigation methods 1 Search Box Find function targeted object search with manual confirmation 2 Mouse Thumbnails discrete random 3 Page number increase decrease button dis crete sequential 4 Mouse Scrollbar Scrolling to the next page by click continuous sequential Hypothesis The task completion time gets longer from the navigation method 1 to 4 Documents for test 1 document is selected for the practice test and 3 documents are for the formal tests In order to analyze the performance for navigating documents of different length and to average the task completion time the length of the three documents selected for the formal tests are 10 pages 44 pages and 106 pages And the target is on the 5th 23rd and 66th page respectively Procedure For each of the four navigation methods the users is asked to do navigation tasks on the three selected documents The order of the four navigation methods and the order of the three documents for each navigation method will be pseudo randomized so as to counter balance the learning effect of the user Before the formal tests of
14. A 11 Evaluation block diagram of search box find function for digital document A 12 Evaluation block diagram of page number input box for digital document 122 127 128 129 130 131 132 List of Figures A 13 Evaluation block diagram of page number B 1 B 2 B 3 BA B 5 B 6 B 7 B 8 C 1 C2 C3 increase decrease button for digital document 134 Evaluation block diagram of mouse stylus touch screen scrollbar for picture interac five TAP aie sk dd deme ao dos s 136 Evaluation block diagram of mouse stylus touch screen tap and drag for picture in t ractive map DEEN 137 Evaluation block diagram of mouse stylus touch screen touch and go for picture in teractive map 138 Evaluation block diagram of mouse stylus touch screen planar for picture interactive MAP oy a rege leger A aee i RES LENE endo 139 Evaluation block diagram of keyboard ar row keys for moving and key for zoom in out for picture interactive map 140 Evaluation block diagram of joystick for pic ture interactivemap 141 Evaluation block diagram of TWEND for picture interactive map 142 Evaluation block diagram of search box for picture interactive map 143 Evaluation block diagram of mouse stylus touch screen timeline slider for digital audio 146 Evaluation block diagram of
15. H 376365 13129659 742012 740151 Figure E 7 Raw test data for digital video 1 169 ered REG AN 2 ee E somos EE el daa an GE E 12 81955 17 1 603 058225 10683845 396712704 3 524494 79043733 13083966 27 2 222037064 S3 709635 376764358 13761714 77994041 ae ana 24676437 62776881 37820714 21308059 90253731 20 648293 as 222 633728 3772578 7731 82 0983142032 3310011 28 2 225000351 13783311 380636383 4142635 10159731 3383328 UO gig seb Hit EE 19 14 227 055a69 35621252 383228577 80798378 62524541 EIE iu uid EE Figure E 8 Raw test data for digital video 2 170 E Raw test data ered e 983168 DEIA Figure E 9 Raw test data for digital video 3 Bibliography Jason Alexander Andy Cockburn Stephen Fitchett Carl Gutwin and Saul Greenberg Revisiting read wear anal ysis design and evaluation of a footprints scrollbar In Proceedings of the 27th international conference on Human factors in computing systems pages 1665 1674 Boston MA USA 2009 ACM Ronald M Baecker and William Buxton Readings in human computer interaction A multidisciplinary approach Morgan Kaufman Publishers Inc Los Altos CA USA January 1987 Patrick Baudisch and Ruth Rosenholtz Halo a technique for visualizing off screen objects In Proceedings of the SIGCHI conference on Human factors in computing systems pages 481 488 Ft Lau
16. No sann A EES e E j j EIL A j pen k TR w j N ba owa ad A E Xx DIM Wi P ed IE m ml Tr AEN prep 133 juoumoop TE3STp 103 xoq mdu 19quinu aged jo ure18erp porq uonen eAq ZU 3 nd q A Evaluation block diagrams for digital document 134 jueurnoop ejiBrp 10 uoynq ISPIMIP aseazour requnu a3ed jo ure18erp porq uonenfeag ET Y 3 nd J ma mwan l ten m i genanv ER 2 4 J o A aa Yi mommmo gt na GINTAN u an pouvwa ao gU Pd dl pen j EER T ag m EE E E n n fed moun en Y e e E 1 L zaa y u pen i wu W ae Nr m RE k A N T Rm ap Sen d o muombe Y mmi n L eh uou 7 wee ke be eg TEE t ADEM Appendix B Evaluation block diagrams for picture interactive map 135 dew aAnJEIBJUI ampid 10j 1eqj 025s USIMS yono sSNJA S asnow Jo ure18erp A90 q UOHENTLAJ r g am IR N A e Pi f n n fre a SE 1 ag MT f mm T s f nd N C wumi S wapa w L mopy d L paag d e E d j B Evaluation block diagrams for picture interactive map 136 137 deur aAoerajur a1njoid 103 Se1p pue dej usams Yona sn 4s asnow jo ure18erp xpo q uonenjeAg Z q andi B Evaluation block diagrams for picture interactive map 138 dew aAnoerajur ampid 101 o8 pue uonoj uea1os yono sn ss asnou jo ure1derp xpo q uorenfeaz g am LI wed patap pa ie
17. The comparison evaluation did in Lee 2007 showed that position based control is better than rate based control for locating a target between 90s and 100s in audio navigation But the authors also believed that there exists a crossover where the rate based control will exceed position based control And in real navigation tasks the target to be searched for will not be in such a short distance like that in the experiment which is shorter than 2 minutes Manipulation The most navigation methods today are tool based That means the user needs to manipulate some widgets to navigate the digital media There are few direct manipulation techniques up to now The well known tech nique for document and picture map navigation is drag and drop And in these years some direct manipulation techniques have appeared for navigating digital video For some tasks tool based navigation can be more powerful than direct manipulation For example if the user needs to browse a video clip direct navigation is not applicable for doing that However the reason why direct manipulation is laid at the right side of tool based navigation is because it is more accurate and more natural The advantages of direct manipulation for digital video have been discussed in 2 2 Navigation designs For interactive map naviga tion it is more flexible than the two dimensional scrolling bar Therefore designing more navigation methods using direct manipulation techniques
18. also ranked significantly lower than the other two while no significant difference was found between the tap and drag and touch n go In addition although the results for tap and drag and touch n go are comparable the participants feedback showed that the touch n go technique is the most preferred on small devices particularly if mobility is important since it re quires less manual effort In contrast the two dimensional scrollbar was the least preferred since the moving horizon tally and vertically can only be done separately and many participants commented it was slow and inconvenient to use Burigat et al 2008 did a study to compare two Zoomable User Interfaces with Overviews ZUIOs a classic ZUIO and a Wireframe ZUIO against a classic Zoomable User Interface ZUI to navigate three types of large informa tion spaces which are maps diagrams and web pages on a relatively small screen which is the mobile device screen in the experiment The results showed that users spent significantly less time to search for targets in maps with Classic ZUIO than with Classic ZUI Classic ZUIO re quired significantly fewer zoom actions than Classic ZUI in both MapTasks and DiagramTasks while it also required significantly fewer pan actions than Classic ZUI in Map Tasks In the SpatialMemoryTask the results showed that users made significantly less errors with Classic ZUIO than with Classic ZUI for maps The authors concluded that th
19. block diagrams for digital audio 145 C Evaluation block diagrams for digital audio 146 Orpne eyi3rp 103 1opi s eur ourn USAS yono sn 1s ssnoui jo ure18erp 190 uonen eAq TJ enSrq f rda DE EN i A ma N Pereeteag i indu j medien eg i Pazi 147 orpne Te8rp 105 ayer xpeq erd Surseoa1our uojing purmar premuoj 1961 USIMS yono SNIAIS ISNOVI jo UIerderp Porq uonen eAq z amS Jn i C mun oc L F X k ga N E ham vaag y perdes k d amp momy d ko ama 7 Wi E Ne A EET M d eg AN mm YA zu ex M E d c Ma A apenn J JM parraat E gt f Ne apt ate 2 7 Pr no gt A A F P d f wio a NN Patel d W os a S G K sse AE Ee m ahy Y aiT i d mogurg 1 amy reg i d M a gas smag VW po b at nen p pambo o san sat psum ERA j EA aC C Evaluation block diagrams for digital audio 148 orpne Ter3rp 103 Sutddrys aurezg uoyng purma PIEMIO 1991 uaa15s yono SNIAIS asnou jo ure18erp poo q uonen eAgq g am ti wo praimer noun pd perlar nme 215 1153 on 3 een o t STORE gt H a gt moputy ruanb a5 ERG MOER sts A on ayangy d AN Pego 4 Schu y weI gr un d i mag 149 4 orm ug f XY o nY orpne ey131p 103 papeo Sunds poum anus jo ure18erp po q uonen eAq p
20. brows ing a digital file locating a certain part of the content and searching for some specific information in the content etc The way people accomplish these tasks is often using an input device to manipulate a kind of navigation control provided by the software Therefore talking about naviga tion method in this work it often refers to the combination use of an input device and a kind of software implemented navigation control Speed efficiency natural mapping and good usability etc have been the criteria to evaluate a navi gation method Hence for the sake of being able to do more efficient and more natural navigation in digital media a va riety of novel input devices and navigation controls are de veloped Input device Input device no longer means only mouse and keyboard Stylus and touch screen are widely used for electronic devices especially for handheld devices like smart phone and PDA Shuttle wheel joy dial are often used to navigate audio and video clips The click wheel which is well known for the use on ipod acts as another in put device which can be used for navigating many digital media types including documents audio and video Joy stick is used to adjust real time camera view by panning and zooming which is also suitable for interactive map Besides some more natural input devices are also being de veloped by researchers An example is TWEND Herken rath et al 2008 a deformable user interface wh
21. confir mation 2 Tap and drag with some cues provided manual search with the help of predefined frames 3 Tap and drag totally manual search Hypothesis The task completion time gets longer from the navigation method 1 to 3 Maps for test 1 city map is selected for the practice test and 3 city maps are for the formal tests The cities for formal tests are nei 7 1 Task amp Setup ther from the countries of the participants nor top tourist cities Therefore the participants do not have enough knowledge of the city so as to be able to avoid the impact on the performance The targets are selected to be 1 2km 3km and 5km from the city center respectively Procedure For each of the three navigation methods the user is asked to do navigation tasks on the three selected city maps The order of the three navigation methods and the order of the three city maps for each navigation method will be pseudo randomized so as to counter balance the learning effect of the user Before the formal tests of each navigation method the user will have the chance to practice with the coming navigation method on the practice city map After finishing the prac tice test the user can start to do the formal tests The user is asked to press F1 to load the city map and start the time recording and to double click the found target to locate it to the screen center and press F2 to confirm which stops the time recording as well The progr
22. conventions The whole thesis is written in American English He she is used to describe the unidentified third person The plural we will be used throughout this thesis instead of the singular I even when referring to work that was primarily or solely done by the author Chapter 1 Introduction Digital media has been playing an important role in today s life work and education etc as computer technology de velops It is taking the place of traditional media step by step People read news and books on computer in many cases through Internet using pdf doc html files instead of reading newspaper and printed books They use in teractive maps such as Google map to locate the desired places all over the world obtain surrounding information such as hotels bus stations ask for the driving route be tween two cities and to calculate the distance and esti mated time needed to go there instead of buying printed maps and search on them Lecture videos are recorded to execute remote education or to provide to the students for review Therefore the ways people read get information and even take classes are intrinsically changed Comparing to traditional media digital media can be published and distributed faster and more easily it is more conveniently to be shared carried and archived for a long time hence is possible for making a large amount of information and it is more flexibly to be accessed almost from anywhere co
23. e aki anba n Hm R R si RB MERE a mma C amt la ae be HE EI o EE gn on re e H q4 Tram TT p K TAT zm PL P avi A u H mand EZEK TE E rg p r EE TEE gg s pps gs om bee Ton zg m qme Vu o r X n TE ELE e eng eret og om SY pomo w to eV m se em 3 ar TN v emm Vero Re r 9 eats i Ke mem n ay gone a any my uam bra m pats LII E Ln ng t Dd i 3 yo n IE a ram LABELED q LII LLLI I IL Im cmq qui ap hans pasan quatum oum eum pou y amande A y aae n mem ag o ya pa A gf ema Tu He ed cees pami K vine p oe T Meg AOS A an ss e WEL dl 102 A EM 1 b E i mate anh fe rech See Tra a Do ri m c Bete Coman foul Tr antun fete Pa mas Pee Tra e rn KX E te ng snas boe ta manyen 7 Validation Figure 7 5 Test program screenshot search box in interactive map 103 7 3 User test program deur aAoerajur ut Se1p pue dej Joysusams WEIBOLA 183 9 Z IMSI Ed mwn qnam maman mi P erg r on yo geg m tom gt mo amp Pn m ma cma depenn mn vtm gut r get mom D y mde fer sy a fe KA tas vg re IL al ei A i VEe fei Caf EE gn e Pei per tat T 7 Validation 104 deur aAnoeroQqur ut and YAM Se1p pue dej Joysusams We1301d 459 44 AMBY S M mes Ka ayt MEE party aia Hin E mn aut gh centi ap un Ee mh bannan as ban asp quos arb ww r anas R gt Ze dej fe m I
24. indicate different properties of the input devices are connected by three kinds of lines which stand for three composition op Card s design space 10 Other work on design space 3 Related work erators A black line stands for a merge composition e g mouse X axis movement and Y axis movement a dashed line stands for a layout composition e g mouse movement and button clicking and a double line stands for connect composition which works when the output domain of one input device is mapped onto the input domain of another one e g mouse output is mapped to the cursor on screen The number in the circle represents the number of identical devices By putting the circles representing all the existing devices into the design space it can be easily seen where rooms for improvements are Besides the work explored the use of device footprint and Fitts s Law as a test 2 a 5 L 5 a L v gt 5 Force Delta Force 9 100 imiji 10 100 Im 10 100 Ind 1 gt 100 b f1 10 100 be Measure Measure Measure Measure Measure Measure Figure 3 1 Card s design space Some other researchers also attempted to classify the input devices for exploring the design space but did not define a notion of completeness of the design space Foley et al 1984 classified input devices for using in computer graph ics subtasks while Buxton 1983 and Baecker and Buxton 1987 classified only continuous devices according
25. me dia either in an increasing sequence or a decreasing se quence It could be continuous or discrete but the order of the contents accessed should be the same as the order they exist in the media On the opposite side random ac cess is to say that the user can jump to any accessible po sition or frame as desired and does not need to follow the sequence the information appears Furthermore sequen 27 The fourth dimension Access 28 The fifth dimension Navigation Search 4 Navigation of digital media tial access is a subset of the random access Any navigation method which is able to do random access is also be able to do sequential access For example using page up page down key to navigate document is sequential access since the user can only access the next page previous page by pressing the page down page up key while using thumb nails to navigate can accomplish random access since the user can click any thumbnail to go to the corresponding page as they wish regardless the page sequence And fur thermore using thumbnails can also accomplish sequential access by clicking the thumbnails one by one following the sequence of the pages Since a navigation method can hold one property from the first group and one from the second group the access could be continuous amp random continuous amp sequential discrete amp random and discrete amp sequential An example for each of the four kinds of access
26. not something like an entity which can be manipulated But by providing novel representation formats of digital audio may also make di rect manipulation possible For example digital audio can be represented by a sequence of notes or a sequence of waves In this format it may be able to be manipulated directly for some professional purposes 73 5 2 A comprehensive look at the navigation methods in digital media orpne reji3rp 103 o eds UBISIq ge 91314 Aaa a EN i CU NM NE NN a KE 74 5 Design space Notations 1 Mouse Stylus Touch Screen Timeline slider 2 Mouse Stylus Touch Screen Fast forward Rewind button increasing playback rate 3 Mouse Stylus Touch Screen Fast forward Rewind button frame skipping 4 Shuttle wheel Spring loaded 5 Jog dial Click wheel 5 2 A comprehensive look at the navigation methods in digital media 75 5 2 4 Digital video Current design overview By examining the design space for digital audio Figure Current design 5 9 findings obtained from observing the current naviga situation of tion method designs include navigation in digital video There are a variety of navigation methods which is tool based There are some navigation methods which provide prede fined frames There are few navigation methods with direct manipula tion technique Future design suggestions There are already a variety of tool based navigation meth Future de
27. o amS ea Ko kart UE pare wera Vo ZO X xy C en C Evaluation block diagrams for digital audio 150 Orpne eji3tp 10j 29A YAN Terp Sof jo ure18erp xpo q VORENIEAR 6J e1nSrq X _ ie A mopord N E V ON De faa aig pros mere j a ko m aE a Www M ENE m AA f mdu s prt anpray nm ggf and mpg ON d ama A x rr D mamb aT Ba REY d j An 7 R df easi JONN tans E dir Appendix D Evaluation block diagrams for digital video 151 D Evaluation block diagrams for digital video 152 OSPIA Te Zip 10 1apris eur ourn USAS yono SHJ IS asnow jo weiserp xpo q UOHENTLAJ T A IMSI p mpa x MA s N ende weno y J S Dat ka A ai T a E k DE A f ann y smt u n szuieIy Pere greaL m a ae ere S shu wuem ueg HEHE Ad d say 4 j reag merdas A B f 153 OSPIA TeziBrp 103 ounoun padIem a A1s akaysy usamos yono sn 41s asnour Jo ure18erp xpo q uonenjeAg ZO 31314 Sei r mopurg ba S N d AE ME uw As 7 Ie JA e d oes f N f te ET n AS ect e a ko wg FI arpa BT geg B m d be woz AM a uenig sno em pain a Bs aii mm E TR man f pue a E wopurg ur MEN pan es wg W aae Y lint e u L nom 7 pee a NE J vng m N e KAD AT E SE ba e D Evaluation block diagrams for digital video 154 OSPIA e
28. scrollbar was defined as a kind of continuous amp discrete amp random access This is be cause beside the continuous sequential access which can be accomplished by dragging the thumb in the scrollbar it can also be set to jump here when clicking somewhere on the scrollbar The latter one can achieve discrete and random access But in the user test what is expected from the scrollbar is only the property of continuous sequential access Therefore in the experiment setup the system is set to smoothly scroll to the next page when clicking some 91 92 Task 8 Setup for interactive map 7 Validation where on the scrollbar By this setting the user can only jump to anywhere he she wants and thus it mimics the continuous sequential access Document display One page of the document will fill the size of the screen so that user can read all the things on one page without scrolling This makes it possible to forbid the use of other navigation methods when using thumbnails and page number increase decrease button to perform the tasks Test environment consistency The program window will be displayed at the same position on the screen and of the same size without the possibility of change so that the test environment for all users will be the same Part 2 Interactive Map Task Find a required place e g Ardwick Green Park in a city map Selected navigation methods 1 Search Box targeted object search with manual
29. some predefined frames to give the user some information about the con tents so that the user can have some clues or hints to find the target more quickly Using predefined frames is a very helpful technique for accelerating the navigation speed We will discuss it in more details later in this work For ex ample fish eye style timeline slider is a kind of navigation method with predefined frames With the small screen shots the user can get an idea of which part of the video is about which topic And this information will make user locate to the desired part easier Totally manual search which requires the most efforts from the user means the search task is totally done by the user s manual search The navigation method only plays the role for manipulating the digital media but it does not help to find the target at all For example traditional scroll bar and timeline slider both belong to this group since the only thing the user can do with the navigation methods is dragging the thumb in the scrollbar to scroll the document or dragging the thumb in the timeline slider to browse the video And users need to read or browse the whole media content to search for the target While other navigation properties are represented by the corresponding squares in the design space the four effort 51 52 Notations and representation of design space 5 Design space indicators will be represented by some extra signs filled in the
30. space and having a comprehensive look at them the navigation methods are represented by numbers in the design space Similar with the represen tation of effort indicators added in the circles which is introduced in 5 1 3 Notations and representation no extra sign means direct positioning plus sign repre sents confirmation needed small circle sign represents with the help of predefined frames and cross sign rep resents totally manual search The difference is the signs are added after the numbers here 5 2 A comprehensive look at the navigation methods in digital media 5 2 1 Digital document Current design overview By examining the design space for digital document Figure 5 6 findings obtained from observing the current naviga tion method designs include Most navigation methods are tool based and position based control For sequential access there are fewer navigation methods with rate based control than that with position based con trol There are very few navigation methods which use direct manipulation techniques Most navigation methods with continuous access can do sequential access only Almost all targeted object position search methods require user s manual visual search effort Future design suggestions For any digital media the goal of a good navigation method is to correctly locate the desired information in the digital media as fast as possible For digital documents th
31. there is an approximate linear relationship between the task completion time using any of the first three navigation methods and the target page num ber there is no linear relationship shown between the task completion task using search box and the target page num ber Significant difference is found for all of the first three navigation methods to show that navigating document 3 is significantly slower than navigating document 2 and nav igating document 2 is significantly slower than navigating document 1 p 0 01 for all It is found that there are several very large numbers in the row data which are weird Per observation during the user test the reason may be misunderstanding the task require ments Especially some participants misunderstood the task description because of the language since many par ticipants are not English native speaker and some of the participants are from Asia whose native language struc ture is different from that of English Therefore an extra 75 Analysis and conclusions Figure 7 12 Relationship between task completion time and target page number analysis was done by using the numbers that are in the range of average 2 standard derivation which waives the weird data and uses the remaining 95 of the original data set according to the normal distribution The analy sis result shows that by waiving the weird data significant difference is shown between the average task completion time by
32. to the scene when video time is at 5 minutes is targeted playback position search while navigating to the playback position when the first runner crosses the line is targeted object position search Especially coordinate position is used for digital docu ment and picture interactive map and playback position is used for digital audio and digital video The coordinate position also includes the meaning of targeted page num bers And the object position for audio means a desired note or the start of a desired melody etc 42 Navigation methods Having the taxonomy of navigation the navigation meth ods can be grounded into families and thus to be analyzed in abstraction according to their properties In the following section all the major existing navigation methods for the four digital media types addressed in this work will be listed They are the ones which will be put into the design space and analyzed in later chapters Since a navigation method is usually the combined use of an in put device and a software implemented navigation control as mentioned before the navigation methods listed in the following section are mostly in the form of an input device plus a navigation control Besides the classification will be given for every kind of listed navigation methods accord ing to the taxonomy described above Since most naviga tion methods can do all the three kinds of navigation de scribed in Navigation Search dimension only
33. to the states in the third stage or in the final stage They describe that with which level of 6 2 Notations and representation ure1gerp xpo q Uonen eAa JO suonejoN z 9 SMS oui mpesa mzn haraa Haaess TE ATIS L sawe porgapead po dq Aa BOOT EEN Soos pnmo PARAN AL Eumornsod pang anno pasee uotiod e say Fu D 84 Steps to generate evaluation block diagram from design space 6 Evaluation block diagram effort the user is able to accomplish which kinds of naviga tion purposes In addition the last type of line is a double line with an ar row which is there in the initial representation of the evalu ation block diagram This type of line is called consequent leading line and it represents the necessary result In the block diagram the line connecting targeted coordinate playback position and targeted object position indicates that if the navigation is able to reach targeted coordinate playback position it must be able to reach targeted object position since every object position corresponds to a co ordinate or playback position Therefore locating the cor responding coordinate or playback position is locating the object position Similarly the line connecting continuous and arbitrary browsing indicates that if the navigation is able to accomplish continuous access it must be able to enable the user to do arbitrary browsing since continuous accessing means that th
34. using scrollbar and that by using page number increase decrease button This result better supports the hypothesis In conclusion although significant difference is not found between all the two pairs of the four navigation methods each of the three documents separately and the average data show the decreasing trend along the four navigation methods except for the search box in document 1 And from the analysis result by waiving the weird data it can be believed that the original results can be improved by in volving a larger number of participants to reduce the effect of the weird data Hence it is concluded that the hypoth esis is proved Therefore we can say that for digital docu ment targeted object search with manual confirmation per forms better than the navigation methods which require manual search random access has advantage against se quential access and discrete access has advantage against continuous access 111 112 Analysis of the user test results for interactive map 7 Validation 7 5 2 Interactive map As shown in Figure 7 13 for all the three city maps search box is faster than tap and drag with cues and tap and drag with cues is faster than tap and drag For map 1 and map 2 which are the maps of Manchester UK and Toulouse France and for the average data shown in Figure 7 14 sig nificant difference is found between every two of the three navigation methods p 0 01 for all And for map 3
35. which is the map of Washington D C U S significant difference is also found between tap and drag and tap and drag with cues p 0 05 and between any other pairs of the naviga tion methods p 0 01 earch bow Figure 7 13 Performance of navigation methods for the three maps It is observed that while map 2 requires the most naviga tion time when using tap and drag and search box map 3 requires the most time when using tap and drag with cues The reason may be most of the cues provided in map 3 are off screen while almost all the cues provided in map 1 and map 2 are displayed in the initial view when the map is loaded Hence users spent more time to search for the off screen cues when navigating map 3 Furthermore Figure 7 15 shows that when the cues are in the current display the task completion time using tap and 75 Analysis and conclusions HT map avg Figure 7 14 Average performance of navigation methods for interactive map drag with cues has an approximate linear relationship with the number of cues provided The difference between the task completion time using tap and drag with cues in map 1 and in map 2 is significant p lt 0 01 This suggests that the task completion time is decided by the number of cues the user needs to check Therefore it implies that the sys tem should only provide the important cues for the user If too many cues are provided the task completion time will also increase In c
36. 1 166 E Raw test data sensezri 1583257 is i52804 zen 13428263 9056556 349837 6 amp 126648 33 672391 714431 214676651 Ges 105472032 R imei EE etri El sa EE EE Ee 38941909 322304191 19 154875 11778299 bei Gea aar an enf ok ano ol an Hen esa Meen 447539447 32412092 16 049419 103565245 40357052 21218862 7762054 Gage 22425109 137090424 31 92005 181 33 21072 69259031 50315552 12 16722070 107083323 279601 a1277393 96313730 37545825 14308739 2473a6204 1352047 les sai 3747192 3613179 35701557 1460726243 22714751 57091434 255362635 19 1 657722008 32192226 13001157 16350531 24022183 12370196 623 390076 87930661 16069332 ST rasne nand l T T ansore zx na Figure E 5 Raw test data for interactive map 2 167 mee plom ty iB esse equ pnm EE a 92326111 10730825 13273441 227380615 31579082 peal sss arw E Figure E 6 Raw test data for interactive map 3 168 E Raw test data Ra ae ET EE L 79043035 7982382 za 223 5979 8415084 R an ama sed rasaae Soe 7 3 244420378 13792879 3196399214 Nut ki usum Germ jmismi rmun misi eun INN E sa column sane E 8 3 237 12957 6300067 3756225178 32213880 78113747 3 990074 Si mui Seed ed up CINE 27007 92 222789062 lei EE 9 3 226500092 23545797 3755475308 941559 58792934 3100561 mos 223033303
37. 7 Unsupervised test process Because the user test can be processed in an unsupervised way it is possible to do several tests simultaneously with several users Therefore a large number of participants are possible to be involved in the user test 98 7 Validation Tank 1 Find the second figure in the paper For thes tank you can only uve next page grevicus page button to navigaie Press FI to load the document and start time recording presa FI after you find the tanpet fomen o rg A NZ P of dd e pet d mn men Kr vr Came gd amp GEE etn fr ii dr oem an WA dra ET kou aa e u EE ker Ke tien rd Fa XE rr cm he m par de ono cdit RA pue cm amem coo warm n mp rpm pan s de penna mba a a X oam LL Ka Cogo P ardh L mar TED RT Time record ng started Map Based Horizontal Navigation In Educational Hypertext Pee B Raka y Riccardo Pure Tare sd v UV reier K rr wr ter Pere Parani Conan Larry of Paran wde n Fa derd sai T er Te Tr pes Pehar FA 11368 win Ugo La Mete 15290648 Para tak 1 412 HA K JA De ARRE peterb 9 sis pitt sa szrod fcf pa on t ABSTRACT The TH em vante mont Fr fe quads of men te ers py brg rg omma T For shee vnam grat ei u bereg an mal o Zeit 1 Watt 0 Tue NO en qmm pen fr oem Newel nami o Tut cerra A ged manier of cm eg g tt mm em Ded de hand dy salem eg Z aal Taare af pm and wieso hand bwen owed tied 0 te ken sf lets des X Ek mew Ire chars easi te pestered Te
38. LSE PE TEDES EOSGES TT vee 22616 TE BREED ETALEEZ dETEG IF Liese EDEOUZ EE POTWEBT BTYOIFLT Tee kiccca Jess LEGTELS Ire SEET Jost est ertere st ess6sZE ees szetzstr rszsze re osgcr v9oest t lsreusse losies Ice DA eSTIZI SS escroest eters st LAT evESS9LS ASESST Z sssotww LOTLLES est Ire Jesse 3 27 ESOSTO CE Ise EreOROLE GODEYEE Lance rercrtv ir LIT 20960871 GHH LO TH ZOETEGE Ire ICH 29540516 setsteoe ce ect esottgs sztrottr jseeersz ersoegz eserzsor kees cesaret eststyot jeu ser retsasye evilSST era Iesst asgcrs eerste er Iesse er ac sesgos Iesse vu BEE LOL Vreergcz org ESTEPE CELA LEEST EZ parrot Ol Joar IZPIS rot LEGLETS Ir St Josee era etotad6r 16sevr vstess et rer Teresc jeoesst T ker ger rMSST S Ice Sir sTtevw TE eBsYEL3 esesasog retECSoT ert sorevot Jarre reerot tt let eiert raros Ire BEZEOT ET Kate LTH r crog ov Sa it r ETT BOTELS HE Ire DEGSES kee eps es erst EE HE uc psa uang yann mM kl e kos a meer o mani l Kal 165 D 4 magi d Wm el tap and drag tap and drag Bere RS A ORN ER So AM AT GR al EES AM 11 33292002 19220330 17390209 7001133972 124005085 1011a223 22224435 26377001 5393942 1422534 33115619 22049269 1 2 26293434 10 640059 MELLE BS ud rue ELE 1101000 7005500 130336842 14033031 T H TH E H 11898041 E T TT T E L Figure E 4 Raw test data for Interactive map
39. R Apa u NDA WERS X 18 1 puna aq 21 papurbas Y 190401 Jyt opiji pasrbas an 29 not cr Gud Ay sIPIGOr api vul od 30 X ma Guy 63 peu UDP 64 N BUIG um piy JO voran BUL MOS usw ADT Mt pug METING pure nore Mu HE EI RULE LA Op 7 4 Results 74 Results The results of the user test can be found in the following tables for digital document interactive map and digital video respectively Overall speaking the average task com pletion time proved the hypothesis in 7 1 Task amp Setup For digital document the average task completion time of using scrollbar is 34 658262s which is the slowest Page number increase decrease button and thumbnails fol low by the average completion time of 28 991567s and 25 495950s respectively The fastest navigation method is search box by using which the average task completion time is 15 308935s For interactive map the average task completion time of using tap and drag which depends totally on the user s manual search is very long 223 033219s By using tap and drag with cues the average completion time becomes much shorter which is 95 765255s Similar with the results of digital document search box proves to be the fastest way by using which the average task completion time is 23 057424s For digital video the result of using fast forward rewind button increasing speed and that of using timeline slider differs a lot While the average task completion time of using fast forward re
40. RWTH Design space for navigation in digital media Master Thesis at the Media Computing Group Prof Dr Jan Borchers Computer Science Department 2 RWTH Aachen University Xiaojun Ying Thesis advisor Prof Dr Jan Borchers Second examiner Prof Dr Ulrik Schr der Registration date May 20th 2009 Submission date Nov 20th 2009 I hereby declare that I have created this work completely on my own and used no other sources or tools than the ones listed and that I have marked any citations accordingly Hiermit versichere ich dass ich die vorliegende Arbeit selbst ndig verfasst und keine anderen als die angegebe nen Quellen und Hilfsmittel benutzt sowie Zitate kenntlich gemacht habe Aachen N ovember 2009 Xiaojun Ying iii Contents Abstract Uberblick Acknowledgements Conventions 1 Introduction 2 Background 2 1 Digital media 2 2 Navigation designs 2 3 Design Space c eur ce Aere are MR po n 24 Evaluation block diagram 3 Related work 3 1 Designspace 3 2 Navigation method Xix xxi xxiii XXV vi 3 3 Contents User study and evaluation 4 Navigation of digital media 4 1 4 2 Dimensions and taxonomy of navigation Navigation methods 421 Digitaldocument 4 2 2 Picture Interactive map 4 2 3 Digital audio 424 Digital video
41. S this in 90 0 1 10 0 114 1 0 04 00 1 0 05 116 0 Ek 1 0 09471 ih 44 16 NR eT gt 01 Pre Ore amend NAA Vun AL De NL L 1 0 1 2109403 L PT HA LA L 371 EN RA IR Male NA el om 0 Towns 6 17 Uniform terminaticn af VASS is in Na Tio Dy Theres 5 9 a VANN la ant ariba rec tenniast ng if and only if dn be ayi IN ap En a ei a soci license part of 2 we can outakder in turn eech Weer part in n The full wt internects Y if and enly U e beast one of its Dites parto dies Thus ell we toed is a method for docking whether a Enear par of 2 pera wad Nt Let Y er ph md v U bea linear part dt In intersects N E end andy if there esos n dehel Ze L This cen be saved in ung sexe Integer near pragrammiag wh nem O Re VASE and Petri mete are equivalent this also shows that unter ter monat ion af Pete nets in decolebie With gelag rach De cquivaleser a direct nod errer proof cen be made f r Preri ets Such a pevod san be found m Moyen 13001 Theverm 66 page AI Brampie 3 LR Coneider again the vwa VASS of Pigure 7 The sot of vnga af man eengty cycles of the fire VAAS corresponde te the sem linear part 1D 0 1 1 8 111 2 8 640 0490 1 0 00 0 0 1 1 0 011 ACT rem ttem om Com gent tal lya YA N A Arie TR Ibo boe Ai MEN Figure 7 3 Test program screenshot scrollbar in digital document 101 7 3 User test program juoumoop ejiBrp UI spreuquny3 30gsuo222s UIEISOId 459 p 4 31314 qoc
42. USIP 107 spreoq 103s orureu p 10 2nejs US yono sn 41s asnou jo ure1derp porq UOonen eAq CO MB Me ee od O pura A f 313128 d f tutus uri SN d AA a gt iiaa GU n MEE X T sso nunca 3 E f man ware 7 Nu t A am qv patwa nan f an A K WE wameg NON d wo gegen Beside pain d T po 3 AAA X U Pap NEG AG d man f rn EEE N A mopurg ar awpa d A pan weg Qu Fe gt iE L ra r mpm Y L pi 1 atann j Fr orat d Se Vo C ei a Sit 155 OSPIA e LSIP 103 aye1 onen eAq T A andi Aerd 3 our uojjnq PUIMII premioj SE USILOS yono sn is asnouw jo ure18erp porq uonen xpeqKerd Surseaxour I E We 7 wepe ON E Ra wi x n Y fi Ze d 4 1 m k a ER i a E e lt lt gt f min N rone asc A ME E os y L Bomm J mamat a mmn os at c AST A i ve ao e ei morera S K ampi es nna A E epad Ly T ez mms 3 pem j A d ei D Evaluation block diagrams for digital video 156 O3PIA e ISIP 103 Sutddrys aurezg uoynq puma premioj 1991 USAS yono SnIAIS asnow jo ure18erp YoTq uonen eAq q am a aopurg en m Mionna momes gt uwenory N Wi em D d gt stonunze y nan nun N gt ae aine qu part qo Va d i qo yi N 1 j pe past i E a j 3 N mem L gu tye A nan y wg N wopu
43. all the con tents in the media users may miss the information they want Therefore in the case if both access methods can make users be able to access all the information discrete access then is regarded to be a more efficient way than con tinuous access to navigate digital media Thus discrete ac cess is laid at the right side of continuous access A user study has been done in this work to prove this hypothesis The results will be presented and analyzed in the validation part Random access is laid at the upper part of sequential ac cess Navigating by sequential access can only access in the sequence the information contents are organized even though the user knows where the desired position or frame is The contents in between the current position or frame and the desired position or frame cannot be skipped any way On the opposite by random access the user has full 5 1 Design space flexibility to locate to any position or frame that is acces sible regardless of the sequence Furthermore sequential access is a subset of random access because any navigation method which enables random access can be used to access the information sequentially since the user has the full flex ibility to decide where to go and thus he she can access in the sequence of the contents Also a user study has been done in this work to prove this hypothesis The results will be presented and analyzed in the validation part Navigation Search Fr
44. am will check the correct ness of the found target by checking the latitude longitude of the found target Only if the target is correct is the user able to proceed to the next task Otherwise the user is asked to continue with the current task and the time record ing will not stop Special requirements Tap and drag with some cues provided The cues are provided in the form of some bubbles indi cating some places or a highlighted line indicating a metro route The information about the distance from the target to one of the predefined frames is provided Examples are 10 minutes walk to the Ardwick railway station and 6 minutes walk to Metro Line A Arenes station Test environment consistency The program window will be displayed at the same position on the screen and of the same size without the 93 94 Task 8 Setup for digital video 7 Validation possibility of change so that the test environment for all users will be the same Part 3 Digital video Lecture video Task Find the time point where a required slide is shown Selected navigation methods 1 Mouse Timeline slider discrete random 2 Fast forward Rewind button increasing playback rate continuous sequential Hypothesis The task completion time using timeline slider is shorter than that using fast forward rewind button Video for test 1 piece of video is selected for the practice test and 3 pieces of video are for the formal tests In
45. an fast forward rewind button rate control in average p 0 01 Besides for fast forward rewind button rate control it shows to be an approximate linear relationship between the task completion time and target length shown in Fig ure 7 18 The difference between the task completion time for all the three pieces of video is significant p 0 01 for all And deviation analysis shows that when using fast for ward rewind button rate control to navigate there is very little deviation of the task completion time This implies that the navigation depends on the system itself to a large extent Since it requires very little user s interaction with 75 Analysis and conclusions 115 4900009000 150000000 309 900000 ond 350 000000 Bu 100 0000 ode 1 10000 even 100 200000 50005000 0000005 FAN END ran con oi tere ine siger Figure 7 16 Performance of navigation methods for the three pieces of video 300 000200 video avg 250 000000 200 000000 120 000009 E T 100 000009 59 000009 Frwy Rw rate control meine dider 90090000 Figure 7 17 Average performance of navigation methods for digital video the computer the task completion time does not vary a lot among the users In conclusion the results show that timeline slider is signif icantly faster than fast forward rewind button rate control 116 7 Validation Figure 7 18 Relationship between task completion time and target length
46. ate the suitability for navigation task go to Chapter 5 1 in a document For this navigation task the required state is the final state targeted object position since it is to find some informa tion in the document The optimal path is from the sec ond stage access directly to the final state without going through the third stage since it is preferred if the navi gation method can directly lead users to the desired ob ject position than if the object position can only be reached by changing the coordinate position The standard block diagram is as shown in Figure 6 3 The required state is highlighted and the optimal path is represented by a red line pointing from the second stage to the required state By comparing the evaluation block diagrams of navigation methods for digital document which can be found in A Evaluation block diagrams for digital document with the standard block diagram and by applying the property reference the suitability ranking from high suitability to low suitability of the navigation methods for this task is shown as below Level 1 Search Box Level 2 Mouse Scrollbar Mouse Thumbnail enhanced scrollbar Mouse Thumbnails Page number Input Box Level 3 Keyboard Thumbnails Keyboard page up page down button TWEND Page number increase decrease button Level 4 Mouse wheel rate based Scrolling Level 5 Tap and drag Mouse wheel position based Jog dial Click wheel Scrolling
47. ates 2000 ACM Thorsten Karrer Malte Weiss Eric Lee and Jan Borchers Dragon a direct manipulation interface for frame accurate in scene video navigation In Proceeding of the twenty sixth annual SIGCHI conference on Human factors in computing systems pages 247 250 Florence Italy 2008 ACM Tobias Lauer and Wolfgang H rst Audio based meth ods for navigating and browsing educational multime dia documents In Proceedings of the international workshop on Educational multimedia and multimedia education pages 123 124 Augsburg Bavaria Germany 2007 ACM Eric Lee Towards a quantitative analysis of audio scrolling interfaces In CHI 07 extended abstracts on Human fac tors in computing systems pages 2213 2218 San Jose CA USA 2007 ACM Bonnie MacKay David Dearman Kori Inkpen and Car olyn Watters Walk n scroll a comparison of software based navigation techniques for different levels of mo bility In Proceedings of the 7th international conference on Human computer interaction with mobile devices services pages 183 190 Salzburg Austria 2005 ACM Jock Mackinlay Automating the design of graphical pre sentations of relational information In ACM Transactions on Graphics TOG volume 5 pages 110 141 ACM New York NY USA April 1986 Gonzalo Ramos and Ravin Balakrishnan Fluid interaction techniques for the control and annotation of digital video fluid interaction techniques for the control and anno
48. based control also called speed based control is to use rate based input devices or navigation controls to ad just the speed of sequential navigation The navigation ac tion of using rate based control is only to change and con trol the navigation speed while the navigation itself is done automatically For example using fast forward rewind button to navigate digital audio or digital video is a kind of rate based control since the user only needs to click the buttons for setting the playback rate while the moving of the thumb in the timeline slider is done automatically 3 Manipulation Direct manipulation Tool based manipulation Direct manipulation is a way to navigate by directly ma nipulating the digital media object itself Examples in clude clicking and dragging the digital documents maps or pictures navigating a piece of video by dragging an in scene object to the desired position etc Currently there is no navigation method using direct manipulation for audio contents Instead of directly manipulating the objects in the digital media content tool based navigation is to navigate the me dia by manipulating some kind of widgets or tools specific input devices or special features of input devices etc Ma nipulating scrollbar thumbnails timeline slider etc are all included in this category 41 Dimensions and taxonomy of navigation 4 Access Continuous Access Discrete Access Random Access Sequential Acc
49. cally from the researchers idea which is based on their use experience of the current prac tices Researchers do not have a measure to evaluate whether the new proposed navigation method will perform over the existing ones before they implement it into a soft ware they do not have an idea about which one is bet ter suitable for a certain digital media type than the other ones and they do not have a clear guidance for them to figure out in which aspects the navigation method should be improved Furthermore they may not have had a com prehensive look at the existing navigation methods so that they may feel confused to decide which navigation meth ods should be selected to be implemented in their software to achieve the best performance and to avoid unnecessary redundancy as well Hence a design space is needed to give solutions And these problems are to be addressed in this thesis work with the design space This thesis work will paint a comprehensive picture about navigation in digital media and propose a framework to systematically describe and analyze navigation methods for different types of digital media Afterwards some guidelines and suggestions for future navigation design will be given and a method for evaluating novel naviga tion methods will be proposed Four major kinds of digital media i e digital document picture digital map audio and video will be addressed in the work The work starts with giving a taxonomy
50. can give user more flexible control make user get rid of the two dimensional navigation or enables a more accurate control the tap and drag technique for navigating digital documents does not have these advantages Furthermore it makes the navigation very slow which requires continu ous user actions to drag the document Recently TWEND improves this disadvantage by mimicking the page flip ping as it enables discrete access page by page However direct manipulation can be more helpful if it is semantic Since direct manipulation may enable users to interact directly with the contents the user may be able to use the contents of the current selection or current focus as the information to navigate to the desired parts For exam ple selecting a word may give user the option to locate to its definition in the same document if any selecting a para graph may give user the information about the locations of the paragraphs about the same topic or pointing to a pic ture in the document may give user the option to lead to the related discussion in the document etc 5 2 A comprehensive look at the navigation methods in digital media Provide user the possibility to access randomly together with continuous access Since most of the current navigation methods with continu ous access can only do sequential access it is suggested that the feature of random access should also be implemented Continuous access is a natural way for reading the
51. confidence that the evaluation tools and methodologies proposed in the work are useful And we believe that the other parts of the work are also rational But the validation work will be done in the future The main contribution of the work is that this work is the first to create a design space for navigation in digital media provide a systematic way to analyze the navigation meth ods and to have a comprehensive look at the current nav igation designs It described the navigation in an abstract way by exploring a taxonomy Besides the work also pro posed a useful evaluation tool to measure the capabilities of navigation methods for different navigation tasks And so far the user test proved its correctness 8 2 Future work Since this work is just a start to explore the navigation in digital media and the technologies of digital media are rapidly developing there is still much work can be done in the future First for getting more convincing data we should redo the user tests for those where weird data appeared Besides there is room for the test program to be improved Time out could be set for those really touch tasks e g tap and drag with totally manual search for the interactive map to make the user test more humane Some action logs could 119 Contribution of the work Future work 120 8 Summary and future work be recorded for the analysis of the data For example the logs of user finding the incorrect targ
52. derdale Florida USA 2003 ACM Stefano Burigat Luca Chittaro and Edoardo Parlato Map diagram and web page navigation on mobile de vices the effectiveness of zoomable user interfaces with overviews In Proceedings of the 10th international confer ence on Human computer interaction with mobile devices and services pages 147 156 Amsterdam The Netherlands 2008 ACM William Buxton Lexical and pragmatic considerations of input structures In ACM SIGGRAPH Computer Graph ics volume 17 pages 31 37 ACM New York NY USA January 1983 Sturat K Card Jock D Mackinlay and George G Robert son A morphological analysis of the design space of in put devices In ACM Transactions on Information Systems TOIS volume 9 pages 99 122 ACM New York NY USA April 1991 171 172 Bibliography Kai Yin Cheng Sheng Jie Luo Bing Yu Chen and Hao Hua Chu Smartplayer user centric video fast forwarding In Proceedings of the 27th international confer ence on Human factors in computing systems pages 789 798 Boston MA USA 2009 ACM Pierre Dragicevic Gonzalo Ramos Jacobo Bibliowitcz Derek Nowrouzezahrai Ravin Balakrishnan and Karan Singh Video browsing by direct manipulation In Pro ceeding of the twenty sixth annual SIGCHI conference on Hu man factors in computing systems pages 237 246 Florence Italy 2008 ACM James D Foley Victor L Wallace and Peggy Chan The hu man factors of computer graphic
53. dia 71 5 2 3 Digital audio Current design overview By examining the design space for digital audio Figure Current design 5 8 findings obtained from observing the current naviga situation of tion method designs include navigation in digital audio There are few existing navigation methods available for digital audio overall No navigation method is direct manipulation All targeted object position search methods require user s manual acoustic search effort There are very few navigation methods which can do ran dom access Future design suggestions Currently there are few navigation methods for navigat Future design ing digital audio contents And especially these navigation suggestions for methods cannot give user any help for searching for some navigation in digital targets in an audio clip Thus the biggest problem for dig audio ital audio navigation is the user does not have any infor mation about the audio content without listening The user has to listen carefully to search for desired targets There fore some suggested possible improvements for providing user valuable information about the audio content are pro posed Search in digital audio It sounds that search is not applicable for digital audio since unlike digital document or interactive map there is no words in the audio content But with the development and improvement of speech recognition computer may be able to recognize the audio conte
54. docu ment but with random access provided users will have more flexibility to locate among different parts of the doc ument An example is scrollbar it enables users to access continuously but user can also randomly locate to any posi tion they want by clicking the corresponding position in the scrollbar if the setting of click on the scrollbar is jump here instead of jump to the next page However this setting of the scrollbar introduces modes Reduce user s manual effort for targeted search It can be observed from the design space that all but one tar geted object position search method require totally manual search of the content by the user It means that the navi gation methods do not give any help to the user for finding the desired target To improve this the technique of prede fined frames can be used The navigation methods could provide user some information about the content e g us ing bookmarks indicating positions of chapter change etc to help users navigate more quickly instead of searching for everything visually Besides the navigation methods could log user s navigation history indicate those places where the user visited most often or most recently A good example is the footprint scrollbar introduced in Alexander et al 2009 it remembers the important places for the user so that it can help users locate to the possible interested po sitions more quickly And the small thumbnails can enable
55. document1 162 Raw test data for digital document2 163 Raw test data for digital document3 164 Raw test data for interactive map 1 165 List of Figures E 5 Raw test data for interactive map 2 166 E 6 Raw test data for interactive map 3 167 E 7 Raw test data for digitalvideol 168 E 8 Raw test data for digitalvideo2 169 E 9 Raw test data for digitalvideo3 170 XV xvii List of Tables 7 1 User test results for digital document 107 7 2 Usertestresultsforinteractivemap 108 7 3 User test results for digital video 108 xix Abstract With the development of computer technology digital media such as e books in teractive map digital audio or video is playing a more and more important role in today s life Because digital media is easily to be delivered shared and stored it is replacing the traditional media types step by step so that is becoming the main medium by which people get information At the same time researchers are try ing to develop more efficient and natural navigation methods to navigate digital media Many enhanced designs of the traditional navigation methods have been developed as well as some novel ones But meanwhile designers may feel lost in the large amount of design choices since there is no way to evaluate these naviga tion methods systematically And designers do not have a clear picture about the current state of the ar
56. double circle in the final stage represents the final state of the navigation meth ods Besides several different kinds of lines are used to connect the states and thus to form the navigation paths Connecting line is used to connect the states in the first two groups in the first stage Since there is no consequen tial relationship between the first two groups of properties the line is a normal black line without an arrow By using this line the connection of a manipulation method and an input property is regarded as a combination of the physical properties The two types of lines with arrow which represent position based control and rate based control are to connect the states in the first stage and second stage They describe whether the control is position based or rate based which can also be told from the design space As mentioned before if the access is random the line stops at the edge of the big circle while if the access is only sequential the line goes into the big circle and stops at the edge of the inner circle The four types of lines with arrow which represent direct positioning the same line with that represents position based control confirmation needed with the help of predefined frames and totally manual search corre spond to the four effort indicators described in 5 1 3 Notations and representation They are used to connect the states in the second stage
57. e overviews bring enough benefits for navigating large infor mation spaces and are worthy of using a part of the view area if they highlight relevant semantic information that users can exploit during the search and the structure of the information space does not intrinsically provide appropri ate orientation cues Dragicevic et al 2008 did user study to compare their interactive video player prototype DimP Direct manipula tion Player which enables user to navigate the video by directly dragging objects in scene with a traditional seeker bar The results showed that the speed of navigating us ing DimP was at least 250 of that using seeker bar in case 21 User study of Zoomable User Interface with Overviews User study of DimP 22 User study of digital audio navigation 3 Related work that users were accurate with both techniques although it seemed to be more precise with DimP The error rate by using DimP was also lower than that by using the seeker bar Besides the qualitative results showed that users pre ferred direct dragging than seeker bar because of the reason that direct dragging was easy to use had immediate results with a high level of precision allowed the user to interact with the video elements he she was interested in not just video as a whole etc Lee 2007 did a user study to quantitatively analyze the audio scrolling interfaces to compare rate and position con trol for navigating audio t
58. e WEE IER EE RER EE WH TEE WEE WODENS ZT 163 7 JUSUINDOP EITAIP 10j e3ep 1991 Mey ZA om emt aa jerosecos Ener Less regrese ker Jeer rsorsrir mes Jeer Jeer rw leck COT LTH LSPLGY 690 PIDEV OT EZEPEZ LT 9060565 62 IGIZE SE Ice v cetc OT Ter GATITZ OE 81 est H Faart ooer Jet Jess vorser rererrst jersroeir raresos aer ees cr eme ess so H CR secorsot sreosewe Jee roer reroess sr aroos re vor DG EGOTOSLZ ESITOTIH zoteos oc BTOZZES E r FEOLTT Iesst ec setvIc W Jost BTEELS6 entt amp SE E MI fessevor sc ess esst jaers sorores cs errevei romero jesssose meses warm ear Eat esst Keen oer essor st revcevor boss sestzroe rarr at st revorss vir LHD 20LE56 67 edd ey t cro6 60 THOLEEG reg 0 ZITEZITLZ jester oo T ELGODT J l r8O T ITEIZALI Joer Trap Eeer arr Fass aezple Bazar as D ecers Jess roser cs ER Reess jerssorsr_jasensses eaccerc sererriT H DR mae est Iesst vor ESSODO LT ESDESE YE erortot c GIG DET LSJPENST TITOS TE FLILEZET TEGEZI POSSE LI STwLEQGE Aa IZIGEE E ST G aen Faas ersaat ees fosser KDYERLYST sesezrsr laccorst aert essetrst est aerer O _ serercac_ sceoseee Jerez jasrioren jvsscacer Jas ber roset remser oreraa est LU IFDESE TT TO00EE LZ ecseterlateon Jans OTrLGrOZ LTH EDEZECL EZDLIZE 2019005 Teo E RT moreret ersmost ezasyon souson scacer s asss mess Jee Eer fosesaro ez _ eresense twr ar nes H T Jereoser fesossoz ts erucseTr jar
59. e designer can firstly make the navigation methods represented in the design space Following steps can be done to accomplish this easy task 1 Start with one of the navigation methods Position circles in the corresponding squares which represent the proper ties hold by the navigation method 2 Add effort indicators into the circles 3 Connect the circles which belong to this same navigation method with black lines 53 Steps to represent navigation methods in design space 5 Design space jueurnoop ejiBrp 103 aoeds usIsap jo uonejoN Ze enSrq Mme BEEN Lll i Be se L 55 5 1 Design space deur aAoerajur a1njoid 101 adeds UBISIP jo uonejoN Ce INSI 5 Design space 56 Orpne reji3Tp 107 o eds uSrsop jo UONBION T S o1n diy o bd i T ial Se eia Lm mm 57 5 1 Design space OSPIA e ISIP 10j adeds USISIP JO uonejoN e am ZEE mmm mmm L D o KA 1 o NE We RN AE Dee 58 Analysis guidelines 5 Design space 4 Redo the above 3 steps to represent all the other naviga tion methods 5 Check whether there exists causality between every two navigation methods and add dashed lines with arrow if needed By doing the above steps the total design of navigation methods is represented in the design space and thus can be evaluated by the methodologies this framework provides Since we want to provide as many as navigation features to the user and to provide a m
60. e most navigation tasks executed by the user are brows ing the content of a new document searching for the inter ested or targeted information in the document or locating the interested parts of information in a document which was read before Hence what a good navigation method should do is to help users finish these tasks as fast as pos sible Combining these goals and the finding of current de signs concluded above the following design suggestions are proposed Current design situation of navigation in digital document Future design suggestions for navigation in digital document 61 62 5 Design space Design more navigation methods with rate based control for sequential access As has been discussed in 5 1 2 Design rationale rate based control can make the navigation time for longer doc uments shorter than position based control if the access is sequential Since the length of today s digital documents becomes longer and longer especially for e books imple menting the feature of rate based control can speed up user s browsing speed to some extent Design semantic direct manipulation techniques for digital documents Previously the most known direct manipulation technique for digital document is tap and drag But actually this di rect manipulation technique does not help a lot for navi gating digital documents The reason for this is unlike in teractive map or video in which the direct manipulation
61. e mouse is over the small thumbnail Besides a number can be typed into the Goto mark box to directly locate the desired correspond ing area The evaluation results showed that the mean acquisition time decreased when the time of revisiting in creased and the Footprints scrollbar required shorter mean acquisition time than the traditional scrollbar when nav igating 40 page documents although both interfaces per formed similarly for 10 page documents Herkenrath et al 2008 proposed a new interaction metaphor by using deformable user interface to deliver gesture inputs They built a hardware prototype named IWEND which uses twisting and bending as interac tion gestures to navigate eBooks interactive maps games etc shown in Figure 3 3 and Figure 3 4 Eighteen gestures were defined to naturally map to the different navigation actions in which some mimics the interaction with physi cal objects Bending the prototype into a horizontal wave form which is like flipping through the pages of a soft cover book is used for scrolling bending an edge of the 11 Footprints scrollbar TWEND 12 3 Related work Lei ven em ore e xa REDI we show rre d gn ted d sr as a r diy mwens Mt e GA mat Goto Mark shortcut E w Revisitation thumbnails 7 TY ha pcm HP np Le Ze comes Aen mo MEER ix ac Vas pie 9 Ze Nd mann a fee Age coloured aem BR including static Kr numeric s
62. e user can access any information in the digital media In conclusion with these notations all the properties which can be told from the design space can also be described by the evaluation block diagram And the two mechanisms keep consistent 6 3 Generation from design space Generating the evaluation block diagram for a navigation method from the design space is simple Generally speak ing it is to add connections among the states where neces sary by checking the properties in the design space It can be done by following the below steps which address the connections from left to right along the navigation process 1 Check whether the navigation method is direct manip ulation or tool based navigation and check which kind of input it is Connect the reachable states by connecting line 6 4 Evaluation methodology 2 Check which kind of access the navigation method is able to accomplish and check whether it is position based or rate based Use the corresponding line to connect to the reachable access state from the reachable input state 3 Check which kinds of search the navigation method is able to accomplish and check the effort indicators Use the corresponding lines to connect to the reachable search states from the reachable access state Following the above three steps the possible navigation paths for a navigation method are easily represented in the evaluation block diagram 6 4 Evaluation meth
63. each navigation method the user will have the chance to practice with the coming navigation 7 1 Task amp Setup method on the practice document After finishing the prac tice test the user can start to do the formal tests The user is asked to press F1 to load the document and start the time recording and to press F2 to confirm the target is found which stops the time recording as well The program will check the correctness of the found target by checking the current page Only if the target is correct is the user able to proceed to the next task Otherwise the user is asked to continue with the current task and the time recording will not stop Special requirements Search Box Find function The task description will not give the exact complete key words which are used for searching by the user Instead the tasks are described in a more natural way which is more like the tasks in the user s idea in practical use For exam ple find the second figure in the paper or find the para graph which describes a prototype named Flex Spread sheet in Chapter 5 will be given as a task Thumbnails The thumbnails are laid out in multiple columns instead of one column and most thumbnails are displayed in the initial view after the document is loaded This gives the user more freedom to jump between thumbnails and thus to provide a totally discrete random access of the docu ment Scrollbar In 4 2 1 Digital document
64. efined frames Navigation methods with predefined frames such as fish eye style timeline and storyboards provide users with use ful information about the contents of different parts If the predefined frames is dynamic and can be adjusted by logging the user s navigation experience and preference the navigation methods will be more intelligent Design accurate playback position search Some professional video processing may need accurate playback position search For example directly typing the time value to locate is a kind of targeted playback position search without any user s manual visual effort Make navigation methods more intelligent By implementing some intelligent algorithms the naviga tion methods may become more efficient For example the SmartPlayer introduced in Cheng et al 2009 which uses adaptive fast forwarding is an improved variant of tradi tional fast forwarding technique It increases the navigation efficiency 77 5 2 A comprehensive look at the navigation methods in digital media OSPIA ejiBrp 107 adeds usIsaq 6 6 andi 4225050252255 E ii E a ME n m E E E Li NH B 78 5 Design space Notations 1 Mouse Stylus Touch screen Timeline slider 2 Mouse Stylus Touch screen Fisheye style warped timeline 3 Mouse Stylus Touch screen Static or dynamic story boards 4 Mouse Stylus Touch screen Fast forward Rewind button increasing playback ra
65. en task com pletion time and numberof cues Bar chart performance of navigation meth ods for the three pieces of video Bar chart average performance of naviga tion methods for digital video xi xii 7 18 A 1 A 2 A3 A 4 A 5 A 6 A 7 A 8 A I List of Figures Line chart relationship between task com pletion time and target length Evaluation block diagram of mouse stylus touch screen scrollbar for digital document Evaluation block diagram of mouse stylus touch screen thumbnail enhanced scrollbar for digital document Evaluation block diagram of mouse stylus touch screen thumbnails for digital docu Inent stesse uec ToS vr Renee AIR RA UR Rs Evaluation block diagram of mouse stylus touch screen tap and drag for digital doc ment see Ee oh Gh ES we Ead Evaluation block diagram of mouse wheel position based joy dial click wheel scrolling for digital document Evaluation block diagram of mouse wheel rate based scrolling for digital document Evaluation block diagram of keyboard up down button scrolling for digital docu MENT ates qt is HO EE EE teats Evaluation block diagram of keyboard arrow keys thumbnails for digital document Evaluation block diagram of keyboard page up page down button for digital document A 10 Evaluation block diagram of TWEND for digital document
66. ent the task com pletion time also decreases along the first three navigation methods But the time for search box is longer than those of the other three which is different from the results of docu ment 2 and 3 From the results we can see that while search box is the fastest navigation method for middle length and long length documents it does not have advantage against the other navigation methods and even performs worse than the others The reason for this is although using search box requires a short navigation time it does require a min imum time for the user to input the keywords But with the length of the document increases the time required using search box does not increase We can see from the figure that unlike the other three navigation methods the task completion time for the three documents with differ ent length using search box does not vary a lot For document 1 scrollbar is significantly slower than page number increase decrease button p 0 01 and thumb nails p 0 01 Search box is significantly slower than the other three navigation methods p 0 01 for all Although the number shows page number increase decrease is 75 Analysis and conclusions Figure 7 10 Performance of navigation methods for the three documents slower than thumbnails there is no significant difference between them For document 2 scrollbar is significantly slower than page number increase decrease button p lt 0 05 thu
67. ered to be removed from the implemen tation By the above four guidelines the designer can eliminate the redundant and unnecessary implementation of navigation methods thus to accomplish the goal of providing users a monotonous navigation environment Besides by repre senting all the navigation methods in the design space the designer can also easily figure out which properties have not been included in the current design and thus to 59 pU Notations of effort indicator 5 Design space Implement other navigation methods to compensate the ca pabilities which the current ones do not have The designer can figure out the squares which represent the properties that are not included in the current design but are needed by the application and then to find suit able navigation methods which fit into these squares or to design new navigation methods to compensate these prop erties 5 2 A comprehensive look at the naviga tion methods in digital media In this chapter all the major existing navigation methods for the four types of digital media will be represented in the design space By doing this a comprehensive look will be taken at the state of the art of current navigation method de signs Also the shortages of the current designs and rooms for improvement will be explored And afterwards some future design suggestions will be proposed For including all the navigation methods for a digital me dia type in a design
68. ered to be removed from the implementation 5 1 Design space However if the application is for interactive map for which direct manipulation could be more efficient than tool based navigation then tap and drag should be retained If A and B fit into same squares and have same effort in dicators only the one with better usability need to be con sidered to be implemented If the representations of two navigation methods in the design space are totally the same it implies that the two navigation methods hold the same navigation properties which means the performance of them should be similar In this case the designer could consider choosing only one of them with better usability to implement in the applica tion If A is a subset of B A may not need to be implemented If one navigation method is a subset of the other one it means that the second one holds all the properties the first one holds In this case what could be accomplished by the first one could also probably be accomplished by the sec ond one Therefore the first navigation method could be considered to be removed from the implementation If Boutperforms A in some dimensions and A does not out perform B in any dimension A may not need to be imple mented In this case A does not have any feature which can outper form B from the observation by the design space Therefore if the ways of A and B to navigate digital media are similar A could be consid
69. es for different navigation tasks are different And for some dimensions one property is laid to the right or upper side of the other one is just because it has some obvi ous advantages against the other one or it is suggested in the future design but it doesn t necessarily outperform the other one in all navigation tasks But still the trend from the left lower part to the right upper part can be very useful to quickly and roughly evaluate the navigation methods The following paragraphs are going to discuss about the rationale for the layout of each dimension Input This is the only dimension in which there is no pre ferred property Absolute input and relative input are the physical properties determined by the input devices or in put controls It just describes how the input works when used to navigate Hence whether the input is absolute or relative does not have much impact on the navigation per formance Control Since the way of rate based control decides that it can only access the information sequentially because users use rate based control by changing the speed of automatic navigation which is usually sequential scrolling or playing we can only compare it with position based control when the navigation is done by sequential access As discussed previously position based control navigates the digital me dia by changing the position of the control mechanism Hence if the user needs to browse through the whole me dia con
70. ess There are four kinds of information access properties They are further classified into two groups indicated by one star mark and two star mark since they describe the different kinds of properties when accessing digital media contents Every navigation method holds one property from the first group and one from the second group The first group focuses on the continuity of information ac cess Strictly speaking on technical the generation and ac cess of digital media is always discrete since the digital au dio and digital video consists of frames while the picture consists of pixels But in this work when talking about con tinuous access it refers to always accessing the next pos sible position or frame which is available to the user with no accessible information to be skipped Opposite to the continuous access the discrete access means the next ac cessed information is not always the next possible informa tion which is accessible for the user The access may skip some parts of the information and directly jump to a po sition or frame For example dragging the thumb in the slider to navigate the document is continuous access while navigating by selecting the thumbnails is discrete access since it always jumps to the beginning of the selected page The second group focuses on the sequence of information access Sequential access means the access can only be done sequentially along the position or frame of the digital
71. ets can be recorded so that we may be able to know the reason of the existence of those weird data And since the user test in this work only proved a part of the framework more complete user test should be done in the future to validate the whole work In this work the design space only addressed four types of digital media which are commonly used But in fact there are more types of digital media and some may become the major ones in the future e g website with links 3D envi ronment virtual reality environment etc Hence extend ing the design space to include more types of digital me dia and explore more properties of navigating digital me dia should be addressed in the future Besides in different use environments the performance of navigation methods would vary somewhat Therefore analyzing the naviga tion in different use environment e g for mobile use in shared environment etc could also be done in the future In addition future work may also address the novel navi gation methods such as speech recognition which is not of the form as today s navigation methods Appendix A Evaluation block diagrams for digital document 121 A Evaluation block diagrams for digital document 122 jueumoop E3ISTP 107 req o108 uaa15s yono SNIAIS snou jo ure18erp A90 q uonen eAq TV 9319 E Ban Leem 2 En ld n n N em Y No mennenm maka NV mao EA al pasi p ba SEN lt N A na u
72. ges in navigation tasks The five dimensions will be discussed in the following section Af ter that a taxonomy according to the five dimensions will be given which abstracts the navigation of digital media and is able to categorize navigation methods into groups The proposed design space and evaluation block diagram which will be introduced later are generated based on this taxonomy 41 Dimensions and taxonomy of naviga tion Figure 4 1 shows an initial version of the design space which represents the dimensions and the taxonomy to be discussed in this section It is presented here in order 23 24 The first dimension Input 4 Navigation of digital media to give an visual and intuitive impression of how the taxonomy discussed below is used in the design space The complete design space with more notations will be introduced in the next chapter 1 Input Absolute input Relative input This dimension is similar to the one used in the Card s de sign space of input devices Card et al 1991 The dif ference is the absolute relative property here does not mean the physical property of the input devices Instead it refers to the input property which is held when combing the use of an input device and a navigation control By this dimension absolute input means the position or time point frame of the digital media is uniquely mapped to the position of the input Locating to a specific input po sition at any time wi
73. gram structure 81 Notations of evaluation block diagram 83 An example standard block diagram 88 Test program screenshot page number in crease decrease button in digital document 98 Test program screenshot search box in digi taidocument 99 Test program screenshot scrollbar in digital document bede es 100 List of Figures 7 4 7 5 7 6 7 7 7 8 7 9 7 10 7 11 7 12 7 13 7 14 7 15 7 16 7 17 Test program screenshot thumbnails in dig ital document Test program screenshot search box in in ter active map N ae rac im ate RA Test program screenshot tap and drag in interactive map Test program screenshot tap and drag with cue in interactive map Test program screenshot timeline slider in digital video Test program screenshot fast for ward rewind button rate based control in digital video Bar chart performance of navigation meth ods for the three documents Bar chart average performance of naviga tion methods for digital document Line chart relationship between task com pletion time and target page number Bar chart performance of navigation meth odsforthethre maps Bar chart average performance of naviga tion methods for interactivemap Line chart relationship betwe
74. he amount of saturation on the red color The red color saturation value on the seeker bar is higher where TLslider SmartPlayer 17 18 Next Event Button Play Pause Button Previous Event Button Stop Button Navigation designs of digital audio Mode Playback Speed nformation Time information b Play Tee diy Time EN a ta 165 3 Related work Playback Speed Auto Manual Dashboard 78 y mmm Seeker Bar Cursor Predefined Event List Figure 3 9 Smartplayer the corresponding video segment has a higher amount of motion which will probably lead to the slowdown of play back The predefined events are generated based on some predefined rules For example in surveillance videos the events are defined based on the appearance of pedestrians cars etc while in news reports the events are categorized into different news types such as financial political and sports news The user test showed that by using the SmartPlayer the av erage manual adjustment time decreased from navigating the first video to the fifth video since the player learned more about the user s preference along the five navigation tasks And the average video watching time using Smart Player is shorter than that using traditional player while the comprehension of contents by using SmartPlayer still remained at a high level The discussion suggested that the predefined events served as good hints to adjust the playback speed And since Sma
75. he shading for each square slightly increases from the left part to the right part and from the lower part to the upper part as well This shading deepness change is intro duced to imply that the properties represented by the right upper part of the design space are preferred to the ones rep resented by the left lower part of the design space Hence for every dimension the more preferred properties are laid out to the right or to the upper part of the less preferred properties The reason to organize the design space in this way and to use the shading deepness as a visual implica tion is that it is very easy for the designer to roughly tell how good a design is by looking at in which part of the de sign space the design lays And it tells the designers to try to design the navigation methods which have the proper ties described in the right upper part of the design space However although there exists the implication that the properties in the right upper part are preferred to the ones in the left lower part it is not necessary to say that a navi gation method must outperform another one which lays in the left lower part of it The reason for this is we can only say the properties in the right upper part are preferred in a SSES N M MN 46 Design rationale Design rationale of Input Design rationale of Control 5 Design space large part of the navigation tasks but not all The preferred properti
76. he she sees the required slide either from forward or backward He she does not need to locate the exact time point where it changes to this required slide from the previous one Test environment consistency The program window will be displayed at the same posi tion on the screen and of the same size without the possi bility of change so that the test environment for all users will be the same 7 2 Participants Since four navigation methods for digital document need to be tested there are totally 24 permutations for the order of the four navigation methods Therefore to fully balance the learning effect 24 versions of the test programs are de veloped For each version 3 participants are expected for obtaining an average data 72 participants took part in the experiment with every 3 participants did the same version of the test program The participants are all students in the university 52 of them are male and 20 of them are female The average age of the participants is around 20 They are from Europe Asia and America and study computer science communica tion engineering and economic engineering etc Most of Participants 95 96 24 versions of the test program Benefits of the test program 7 Validation them had previous experience to use the required naviga tion methods in the user test 7 3 User test program The user test is run by the test program As mentioned before since there are totally 24
77. hortcuts D amd SLI edo savis aama e Si a he zu ws tees and w hy Da T TE Ben Ew woy Wing w Ala e me sage de road au Ge Ges Ah bas Ge canat a du eee paten of du Borm Te ju E ded thumbnail kon xpan umbna Bar dohiy ss u sann m ene Ge Zen Fy a fc akne t aoe lu dl KK Lier vyen pnis w fe duncmom s gror Ten t dem nou kon gz avan he Zenn an de arbes d konn un se ON w ka ad w Ta m m o Tai de m wm ant Figure 3 2 Footprint scrollbar prototype is for a single page flipping bending completely along its horizontal axis can be used for zooming in and out when navigating interactive maps etc The angle of bend ing may decide the speed of continuous navigation such as the scroll rate The work also included an experiment to evaluate the efficiency of the gestures Furthermore in a second version of the work shown in an interactive exhibit TWEND serves as a direct manipulation technique to navigate eBooks By projecting the images of the eBook pages onto the TWEND surface it is used as an eBook reader which can be bended to naturally mimic the page flipping action Figure 3 5 3 2 Navigation method 13 Figure 3 3 TWEND QS 1 ETD e SS oS we Figure 3 4 Flipping with TWEND 14 3 Related work Figure 3 5 Reading in TWEND 3 2 Navigation method Baudisch and Rosenholtz 2003 developed Halo a vi sualization technique that shows users the location of off
78. ich uses a block of foam as the input device to navigate by bending it More details about it will be introduced later in 3 Related work Furthermore besides these input methods which are based on some physical devices there are some other ones which are based on advanced computer technologies like speech recognition and pattern recognition They are regarded as the next generation input methods 2 2 Navigation designs Navigation control While creating novel input devices researchers are also developing more efficient software im plemented navigation controls to make navigation tasks easier Scrollbar is widely used for navigating digital docu ments Today some kinds of enhanced scrolling technolo gies have been developed which make scrolling more intel ligent so as to speed up the navigation process Examples include rate based scrolling by which user can control the scroll rate speed dependent automatic zooming Igarashi 2000 by which the view zooms out at a higher scroll rate semantic scrolling which only jumps to the next seman tically meaningful point to the user thumbnail enhanced scrollbar which displays thumbnails of important or vis ited pages on the scrollbar to enable quick locating rapid serial visual presentation Sun and Guimbretiere 2005 which replaces scrolling with page flipping at high scroll rate Traditional timeline slider is used to navigate audio and video clips Now for video navigatio
79. igation path questions like whether the navigation method is capable of accomplishing the task and whether one navigation method will perform better than the others for a task are able to be answered With this tool designers are able to evaluate a new design in an intuitive manner Chapter 3 Related work 3 1 Design space So far there has not been any dedicated research work on creating a comprehensive design space for navigation in digital media but there was several earlier publicized work creating the design space for other disciplines such as in put device graphical design jet engine etc from which the idea of this work comes from The most important work related to the design space con cept in this work is Card et al 1991 The initial idea of creating a design space for navigation in digital media in an abstracted way is also from this work Card et al 1991 developed a framework to systematize input devices The author gave a taxonomy based on physical properties of in put devices and generated a parametrically described de sign space using primitive and compositional movement operators They used a morphological design space analy sis method to have an insight into both the properties of the design space and the development of novel designs As shown in Figure 3 1 the circles indicate that the input device senses the properties the corresponding areas show in which the circles are Different circles which
80. imeline The study included three input devices scroll ring a rate control consisting of spring loaded ring jog dial a position control using a solid dial touch wheel a position control which is touch sensitive The users were asked to locate a target between 90 and 100 seconds from the current playback position The result showed that scroll ring rate control is significantly slower than both the jog dial position control and the touch wheel position control while no significant differ ences were between the two position control devices The author concluded that position based control performs bet ter than rate based control for closer search targets 90 to 100 seconds in the experiment and believed that as the search target distance increases the performance difference will become less significant and there exists a crossover point when rate based control becomes superior And the author took 100 seconds as a minimum upper boundary for this crossover point Chapter 4 Navigation of digital media In 2 2 Navigation designs many types of input devices and software implemented navigation controls have been presented And in 3 Related work some important techniques were described in more details By exploring the physical properties and navigation features of the ex isting navigation methods five dimensions which describe the different aspects of navigation are concluded which also reflect the different sta
81. iques At about 50 lines the 1 line notch accelerated wheel performed best by 100 lines both accelerated wheels performed better than the stan dard wheel and rate control scrolling However by 400 lines the rate control scrolling and 3 line notch accelerated wheel were significantly faster than the standard wheel and 1 line notch accelerated wheel Zhai et al 1997 did an experiment to compare four in put devices standard mouse mouse with a track wheel mouse with joystick and mouse with in keyboard joystick for navigating web documents The results showed that the rate based devices mouse with joystick and mouse with in keyboard joystick were significantly faster than the position based devices standard mouse and mouse with a track wheel while there was no significant difference found between the two rate based devices and the two position based devices MacKay et al 2005 did a field study to compare three navigation techniques scrollbars tap and drag and touch n go on mobile devices The target selection time data showed that the scrollbar technique was significantly slower than both the tap and drag and the touch n go tech nique in all of the three conditions sit stand and walk 3 3 User study and evaluation while there was no significant difference found between the tap and drag and touch n go And by the results of participant rankings of ease of use and preference for nav igation techniques scrollbar was
82. is scrollbar fast forward rewind button which increases the playback rate thumbnails and page up page down key respectively 5 Navigation Search Arbitrary browsing Targeted coordinate playback position search Targeted object position search This dimension addresses the final purpose of the naviga tion task which is either browsing the digital media con tents or searching for a specific target Arbitrary browsing is to browse or review the contents car ried by the digital media have an overview of the informa tion delivered or find something which may be interested to the user It is often needed to know about what a docu ment or a video clip is about Targeted search is further classified into targeted coordinate playback position search and targeted object position search Targeted coordinate playback position search aims at navigating digital document or picture interactive map to a certain coordinate position or to 42 Navigation methods navigate digital audio or digital video to a certain playback position while targeted object position search aims at finding the coordinate playback position where or when a certain object in the digital media content is at a desired position or in a desired state In this case users do not nav igate by locating a coordinate playback position Instead they navigate by checking whether the targeted object has reached the desired position or state For example navigating
83. is suggested 47 Design rationale of Manipulation 48 Design rationale of Access 5 Design space In Karrer et al 2008 the user study showed that DRAGON a direct manipulation video navigation tech nique significantly reduces the task completion time for in scene navigation tasks by an average of 19 42 compared to a standard timeline slider In Dragicevic et al 2008 the evaluation results showed that navigating using DimP an other direct manipulation video navigation technique was at least 250 faster than using seeker bar The error rate by using DimP was also lower than that by using seeker bar Users feedback also showed that direct manipulation is easy to use more natural and has immediate results with a high level of precision In MacKay et al 2005 the tar get selection time data showed that the scrollbar technique tool based control was significantly slower than both the tap and drag and the touch n go technique both are direct manipulation techniques in all of the three conditions sit stand and walk Access Since continuous access needs to go through any available information contained in the digital media the time for going through the uninterested contents thus can not be saved On the contrary discrete access can jump be tween some predefined positions or frames But the prob lem of discrete access is if visiting the predefined positions or frames cannot make users be able to access
84. ll definitely lead to the navigation to the same corresponding position or time point frame in the digital media For example scrollbar is an absolute in put for navigating digital documents since one position of the thumb in the scrollbar is uniquely mapped to a position in a document On the other hand relative input does not have this one to one correspondence Instead it means the amount of change of the position or time duration in digital media is based on the amount of change of the input regardless of the exact input position For example mouse wheel is a relative input for navigating digital documents since the amount of document scrolling is based on the amount of mouse wheel scrolling while there is no relation between the wheel position and document position 25 41 Dimensions and taxonomy of navigation ooeds uSIsap ur urouoxej ay pue suorsuownp au T p amSuy EJEEJEJEJEJEJE3E3ESE3E3E3E3EES OIO TTT TT TTT IT LLL LLL CC 26 The second dimension Control The third dimension Manipulation 4 Navigation of digital media 2 Control Position based control Rate based control Position based control means the navigation is done by changing the position of the navigation controls or input devices For example timeline slider is a kind of position based control for navigation digital audio or digital video since the navigating is through changing the thumb posi tion in the timeline slider Rate
85. ly analyze navigation methods both by the design space and by the evaluation block diagram 4 Navigation of digital media abstracted the navigation in digital media by providing a taxonomy which addresses the five dimensions of the navigation input control ma nipulation access and navigation search Then the ma jor existing navigation methods for the four types of digital media digital document picture interactive map digital audio and digital video are classified using this taxonomy 5 Design space then presented the design space created based on the five dimensions and the taxonomy proposed With the notations the navigation methods were repre sented in the design space And a series of analysis method ologies and guidelines for systematically analyze the navi gation methods with the design space were introduced Af terwards by representing all the major existing navigation methods for the four types of digital media in the design space the work had a comprehensive look at the state of the art of navigation in digital media discussed the current design problems and proposed future design suggestions 6 Evaluation block diagram proposed another evalua tion tool evaluation block diagram It is a kind of block di agram based flow diagram generated from the design space but describes the navigation by states and paths The eval uation block diagram is a useful tool to evaluate the capa bilities of the
86. m puter and Internet are available But in the meanwhile media types are changing the ways people interact with media are also changed With digital media the interaction is between human and machine In put devices such as mouse and keyboard are acting as the mediums instead of human hands to accomplish human Usage and advantage of digital media Problem of navigation design Scope of the work 1 Introduction computer interaction tasks The human actions like flip ping a book page and checking a place on the map are replaced by manipulating the slider and typing the place name in the search box etc Therefore the navigations in the digital media are all about using some input devices to manipulate the possible navigation controls provided by the machine for a certain digital media type Using mouse or keyboard to linearly navigate through a document or an audio or video clip by manipulating the slider bar or timeline slider has been a major navigation method for human computer interaction for a long time But nowadays researchers are exploring new ways for nav igating digital media aiming at finding faster more effi cient and more user friendly navigation methods Some are already widely implemented in the software and ac cepted by the people while some are still in the lab stage But the problem is no matter whether they are popularly used ones or proposed ones the development of these nav igation methods are basi
87. mation Sometimes highlighting too many important positions in the picture or interactive map will make the original view look like a mess which is also not good for searching for the targets And it is always not possible to highlight all kinds ofinformation in the view In this case designers can implement some intelligent dynamic indicating methods by which the information related to the current selection or focus will be dynamically shown in the view For example when clicking a railway station in a map the map could show all the railway stations in this city or when clicking a station of a bus all the stations of this bus and the route of the bus will be shown Use overview to provide semantic information In many cases the small display can only show a small part of a big scale picture or map The user does not have any information about the contents outside of the screen Therefore providing an overview with enough semantic information is a good way to help user have an idea about what exist outside of the current view Burigat et al 2008 did some user studies and found that the overview with se mantic information does bring benefit for navigating large information spaces Provide orientation cues Unlike digital document digital audio and digital video in which the browsing is always from the beginning to the end the user has to choose a direction to browse the infor 68 5 Design space mation in a picture
88. mbnails p 0 01 and search box p 0 01 Search box is signifi cantly faster than the other three p 0 01 for all Although the number shows page number increase decrease is slower than thumbnails there is no significant difference between them For document 3 search box is significantly faster than the other three navigation methods p 0 01 for all AI though the number shows scrollbar is slower than page number increase decrease button and page number in crease decrease button is slower than thumbnails there is no significant difference shown between any two of them This result may be related to the navigation task setup The task for document 3 is Find the paragraph which describes a prototype named Flex Spreadsheet in Chapter 5 As long as the user knows the target is in chapter 5 he she can quickly skip a large part of the document by using any of the first three navigation method Thus the time differ ence is reduced 109 110 7 Validation Figure 7 11 shows the average task completion time by us ing the four navigation methods The figure also shows the decreasing trend along the four navigation methods Sig nificant difference is found between scrollbar and thumb nails p lt 0 05 and between search box and any of the other three navigation methods p 0 01 for all document avg Figure 7 11 Average performance of navigation methods for digital document Figure 7 12 shows that while
89. mouse stylus touch screen fast forward rewind button increasing playback rate for digital audio 147 Evaluation block diagram of mouse stylus touch screen fast forward rewind button frame skipping for digitalaudio 148 xiii xiv C 4 C 5 D J D 2 D 3 DA D 5 D 6 D 7 D 8 E 1 E2 ES EA List of Figures Evaluation block diagram of shuttle wheel spring loaded for digital audio 149 Evaluation block diagram of jog dial click wheel for digital audio 150 Evaluation block diagram of mouse stylus touch screen timeline slider for digital video 152 Evaluation block diagram of mouse stylus touch screen fisheye style warped timeline fordigitalvideo 153 Evaluation block diagram of mouse stylus touch screen static or dynamic storyboards for digital video 154 Evaluation block diagram of mouse stylus touch screen fast forward rewind button increasing playback rate for digital video 155 Evaluation block diagram of mouse stylus touch screen fast forward rewind button frame skipping for digital video 156 Evaluation block diagram of shuttle wheel spring loaded for digital video 157 Evaluation block diagram of jog dial click wheelfordigitalvideo 158 Evaluation block diagram of Dragon for dig ial video te wy E ry ema a 159 Raw test data for digital
90. n fisheye style warped timeline is developed by providing screenshots at pre defined frames to give user an overview of the video clip and let them have the possibility to jump to the frames in which they may be interested Zoomable timeline slider gives user the possibility to re scale the timeline for coarse grained or fine grained navigation purposes Similar ap proaches which provide overview and pre defined frames for discrete navigation include thumbnails for digital doc uments and static or dynamic storyboards for video etc Besides indexNavigation control Direct manipulation these tool based navigation controls another navigation class is direct manipulation While tap and drag touch and go for navigation documents pictures and maps have been more and more familiar to the user direct manipulation of video is a newly developed approach and still in its early stage Dragon is such an example Karrer et al 2008 With Dragon users can apply dragging actions to any object in scene they can precisely drag the object to any position on its moving path which leads to the moving of other objects as well Similar direct manipulation systems include DIMP Dragicevic et al 2008 and Goldman s system Gold man et al 2008 etc Since direct manipulation makes ac curate object positioning much easier it can well support Tool based navigation controls Direct manipulation Concept of design space Concept of evaluatio
91. n search not possible 4 Shuttle wheel Spring loaded Relative input Rate based control Tool based navigation Continuous amp Sequential access 5 Jog dial Click wheel Relative input Position based control Tool based naviga tion Discrete amp Sequential access Enhanced Jog dial Click wheel position based navigation with in telligible audio feedback 37 prenon swen Targeted posten reach ccodeats playback bend A ot 4 Navigation of digital media Figure 4 4 Summary of navigation methods for digital audio 42 Navigation methods 4 24 Digital video 1 Mouse Stylus Touch screen Timeline slider Absolute input Position based control Tool based navi gation Discrete amp Random access Enhanced Mouse Stylus Touch screen Zoomable timeline slider 2 Mouse Stylus Touch screen Fisheye style warped timeline Absolute input Position based control Tool based nav igation Discrete amp Random access targeted playback position search not possible 3 Mouse Stylus Touch screen Static or dynamic story boards Absolute input Position based control Tool based nav igation Discrete amp Random access targeted playback position search not possible 4 Mouse Stylus Touch screen Fast forward Rewind button increasing playback rate Relative input Rate based control Tool based navigation Continuous amp Sequential access 5 Mouse
92. n block diagram 2 Background the tasks like video cutting scientific footage analysis de tailed sport scene review and video annotation etc 2 3 Design space Since a bewildering variety of navigation methods exists for digital media a way to systematizing them appears to be necessary A design space is a kind of collection to or ganize the existing designs in an engineering discipline As proposed in Card et al 1991 with taxonomy it is possible to organize the designs in terms of abstraction by ground ing individual designs into families which gives an insight into the design space classifies existing designs has a com prehensive look at the state of the art identifies design problems and possible improvements and suggests future designs In this work the way of creating the design space is from the idea of Card s design space According to the properties of navigation several dimensions will be abstracted And then taxonomies are made for each dimension Having the taxonomy the navigation is classified at an abstract level and the designs can be grounded into families which lead to a parametrically described design space 2 4 Evaluation block diagram Besides the design space a new tool evaluation block dia gram will be proposed in this work as a part of the frame work The evaluation block diagram is a kind of vari ant of block diagram which describes the navigation pro cess Through analyzing the nav
93. n vcram nly L Kip povwid De be des Ban a KK e remar Ee enye eg ha m r al en ems geng e geit ond frrr bag rm sg rr ge ven te d Mw of o ode air reg ge Mag gr ohm encia ingentem kai e kee m gee ren a payi t mei ge payen us an ege Za Aa sian Wa fried Ur Ton ies d KE POUDI Ei sept Ir ta Gwan o Fee ter v re er aed nares ve vd bes kass isi ad A N a serer oos Kerr Ked 7741 UE vol D CH II jt Pii li Mi ridi De amm gemma mmm Mra tedocha Ge reckon fe anch and Ka de mw re A vd eus var Figure 7 1 Test program screenshot page number increase decrease button in digital document 7 3 User test program 99 Task Find she paragraph woah describes a prototype named Flex Spreadsheet in Chapter Presa Enter to find the target and presa down ap ker on the keyboard to Find the meat previous cocerance of the vearch renult Press FI te lead tha document and start time recording press F2 after you find the target a o Target found mrorredt press f anaie after Anding the correct target Chapter 2 Related work in thes chapter we finet try oo rud a cateyportzatiom for te Thai ww ER Dte t N ide a lut wt amdericandisy of the benefits and defuit m ouest af De respective Ar renn For bemer o th IDE ww etate r ae thet we deeg dor an anm en te support hes rung mine e Single Information pace alow wars dune dann ve Sege ugan meterme peace urra fem omm ngton
94. navigation methods by comparing with stan dard block diagram for a navigation task The generation method was explained in the chapter followed by the eval uation methodology An example was also given to explain how to use the evaluation tool 7 Validation described a user test which aimed to verify the correctness of the analysis methodologies and framework proposed in the work A test program was de veloped to run the user test in a system controlled way As expected the test results proved the hypothesis For digital document it proved that targeted object search with man ual confirmation performs better than the navigation meth ods which require manual search random access has ad vantage against sequential access and discrete access has 82 Future work advantage against continuous access For interactive map it proved that targeted object search with manual confir mation performs better than manual search with the help of predefined frames and the latter performs better than totally manual search And for digital video it proved that discrete random access has advantage against contin uous sequential access These results are consistent with the analysis methodologies provided in the work and are the same as the results derived from the evaluation block diagram Because of the time and resource limitation the user test was only able to verify the correctness of a part of the work However the results gave us
95. nput Position based control Direct manipula tion amp Tool based navigation Continuous amp Discrete amp Se quential access 11 Search Box Find function Relative input Position based control Tool based navi gation Discrete amp Sequential access only targeted object position search is possible 12 Page number Input Box Absolute input Position based control Tool based navi gation Discrete amp Random access 13 Page number increase decrease button Relative input Position based control Tool based naviga tion Discrete amp Sequential access 4 2 Navigation methods U t wq etim juoumoop eji3rp 107 spoujeur uoreSrAeu Jo reurumg ZE JINGI 34 4 Navigation of digital media 4 2 2 Picture Interactive map 1 Mouse Stylus Touch screen Scrollbar Absolute input Position based control Tool based navi gation Continuous amp Discrete amp Random access 2 Mouse Stylus Touch screen Tap and drag Relative input Position based control Direct manipula tion Continuous amp Sequential access 3 Mouse Stylus Touch screen Touch and go Relative input Position based control Direct manipula tion Continuous amp Sequential access targeted coordinate position search not possible 4 Mouse Stylus Touch screen Planar Relative input Position based control Tool based naviga tion Continuous amp Discrete amp Sequential access targeted coordina
96. nt with high accuracy Therefore using typed keywords or keywords by speech input may be possible to accomplish search in the digital 72 5 Design space audio by matching the keywords with the corresponding audio content in the future Besides for music files it may be possible to design some novel musical keyboard based input devices to match the notes in the audio file Design navigation methods with predefined frames tech nique With the audio analysis and speech recognition techniques the navigation methods may provide some predefined frames for the user For example indicating the positions where there are speech changes indicating the different types of contents in the audio or indicating the repeated part of the current selection etc Visualization of audio A reason why users have difficulty to know the audio con tents is the audio is not visual to the user With the au dio analysis technique it may possible to recognize the au dio contents and to automatically generate scripts which can serve as the visual representation to users For exam ple when pointing to a position in the timeline slider the system can provide a certain amount of scripts Thus the user is able to know what the following part of the audio is about without navigating to that position and listen to the audio content Direct manipulation There is no navigation method which enables direct manip ulation at the moment since audio is
97. odology Having generated the evaluation block diagrams for the navigation methods from the design space it is possible to evaluate these navigation methods Generally speak ing the evaluation methodology is comparing the block di agrams of the navigation methods to the standard block diagram for a navigation task to have a ranking of the nav igation methods by the suitability for the task There are two important notions in the standard block dia gram required state and optimal path A standard block diagram is used to describe the requirement of a navigation task Therefore in a standard block diagram one or more required states are indicated to tell the states which must be reached through a navigation path by using a navigation method if it is able to finish the task If there exists a navigation path in the evaluation block diagram of the navigation method which leads to the required state the navigation method is regarded to be capable of accom plishing this task Otherwise it is regarded not being able to finish the navigation task Besides the required state optimal path is another mea sure to evaluate the suitability of a navigation method for a navigation task As long as there is a path reaching the required state the navigated method is regarded to be ca Standard block diagram for evaluation Required states Optimal path 85 86 Evaluation preferences 6 Evaluation block diagram pable f
98. of navigation in digital media based on the several dimensions which de scribe the properties of the navigation tasks The design space is then drawn according to the taxonomy with the notations and analysis methodologies explained Having the design space we will have a comprehensive look at the existing major navigation methods for the four media types addressed in the work The current navigation method practices will be analyzed and summarized and future de sign suggestions will be proposed according to the analysis of navigation properties and current designs which will give us a picture to explore the potential aspects that can be improved Besides newly designed navigation methods can also be added to the design space in the future and the design space can also be used to analyze other novel digital media Based on the design space the work proposes another tool in the framework which is called evaluation block diagram It is a kind of block diagram which is corresponding to the design space and can be generated from the design space The evaluation block diagram is used to describe the nav igation process of a certain kind of navigation task for a type of digital media and accordingly to evaluate whether a navigation method fulfills the requirements of the nav igation task and among a variety of possible navigation methods which is the most suitable one It is useful for helping developers to select suitable navigation method
99. om top to bottom the order of layout are targeted object position search targeted coor dinate playback position search and arbitrary browsing This is consistent with the purpose of navigation The fi nal goal of a navigation task is to find the object position in the media It could be a sentence in the document a place in the map a certain word in an audio clip or a cer tain object in the video being in a desired place Hence if the navigation methods can directly find the targeted ob ject position it is the most preferred Targeted coordinate playback position is the second preferred since finding the correct coordinate playback position is a way leading to the targeted object position and coordinate playback po sition is always what the user needs to adjust when doing a navigation task Arbitrary browsing although commonly used is the least preferred since it never provides a way directly go to a target Users have to locate the target by manual browsing of the contents Blind Spot As mentioned before the rate based control itself determines that it is only possible for sequential ac cess since the user can only change the navigation rate while the navigation itself is automatic scrolling playing etc Therefore the 24 squares in the design space which combine the properties of rate based control and random access are not possible for navigation methods Thus they are considered as invalid squares called blind sp
100. on method which can do random access can also ac complish sequential access Therefore sequential access is represented as an inner circle within the big circle which represents random access both for discrete access and con tinuous access And if the navigation method is able to ac complish random access the connection line will stop at the edge of the big circle while the line will go to the inner cir cle if the navigation method can only accomplish sequential access The third stage represents the navigation effect the access brings It either enables the user to do arbitrary browsing or locate to a targeted coordinate playback position Al though the targeted object position search is also one of the navigation effects it is separated as the fourth stage which 81 6 1 Structure ampnys ure18erp Porq UOHENTLAJ CO SMS ERE 82 Notations and representation of evaluation block diagram 6 Evaluation block diagram is also the final stage since locating the targeted object po sition is always the final purpose of a navigation search task Strictly speaking arbitrary browsing could not be regarded as a goal of navigation This final state is distin guished from the other states by using a double circle 6 2 Notations and representation Figure 6 2 shows the notations used in the evaluation block diagram As mentioned in the previous section the circles represent the navigation states and the
101. onclusion the results show significantly that search box is faster than tap and drag with cues and tap and drag with cues is faster than tap and drag Therefore the hypothesis is proved For interactive map targeted ob ject search with manual confirmation performs better than manual search with the help of predefined frames and the latter performs better than totally manual search This is also the preference given for the effort indicators Furthermore the deviation analysis also shows that tar geted object search with manual confirmation has much less deviation than the other two methods and the devi ation using totally manual search is even bigger than that using manual search with the help of predefined frames 113 114 Analysis of the user test results for digital video 7 Validation Figure 7 15 Relationship between task completion time and number of cues 7 5 3 Digital video As shown in Figure 7 16 for all the three pieces of video timeline slider is significantly faster than fast for ward rewind button rate control p lt 0 01 for all The time difference between the two navigation methods increases when the target length increases But although video 2 requires the most time to navigate when using fast for ward rewind button rate control it does not require the most time when using timeline slider The average data il lustrated in Figure 7 17 also shows that timeline slider is significantly faster th
102. onotonous navigation envi ronment as well which means duplicated navigation meth ods for accomplishing the same navigation effect are not wanted following analysis guidelines are suggested A and B are used to represent two navigation methods here If A points to B and A does not outperform or compensate B in some dimensions A is considered to be redundant with regard to B If one navigation method points to the other one it tells that the manipulation of it will cause the work of the other one That means manipulating the navigation method is to indirectly manipulate the other one and then to have navigation effect on the digital media Therefore if the first navigation method does not provide any feature which out performs the second navigation method the first one has no meaning to be implemented since the second one could accomplish anything the first one could and it can directly have effect on the digital media with no other navigation methods in between For example as shown in Figure 5 2 tap and drag points to scrollbar Tap and drag does not outperform scrollbar in any other dimensions except that it is direct manipulation which compensates the scrollbar But for navigating digital documents direct manipulation has no advantage against tool based navigation in most cases Therefore in the ap plication for which direct manipulation does not have ad vantage against tool based navigation tap and drag could be consid
103. or a task But for some tasks there may be some optimal paths by going through which the navigation will be more efficient This is defined by the optimal path in the standard block diagram The navigation methods which have the paths that are the same as the optimal path are re garded more suitable for the navigation task than the navi gation methods which do not have such paths Furthermore for evaluating the suitability within the nav igation methods which have the optimal path and within those which do not the following preferences of naviga tion properties which are listed from high preference to low preference are referenced The listed properties cor respond to the notations in the evaluation block diagram thus can be easily told from the different types of lines and the states reached Direct positioning Confirmation needed With the help of predefined frames Totally manual search Discrete Random Continuous Random Discrete Sequential Continuous Sequential Rate based control Position based control Since different navigation tasks have different require ments the required state and optimal path for every kind of task are different Therefore the suitability ranking of navi gation methods is only meaningful with regard to a specific kind of navigation task In the following part an example will be showed to explain how the evaluation block dia gram is used 6 4 Evaluation methodology Example Evalu
104. or in a map Hence no orientation in formation will make the searching be in a hopeless tangle Therefore orientation cues such as off screen indicators are good information to tell the user where to go Besides the above proposed techniques for providing searching cues designers could also Design navigation methods with direct manipulation tech nique Since direct manipulation technique can provide users with the flexibility the two dimensional scrollbars cannot and it is more natural direct manipulation is a meaningful tech nique to be implemented for navigating picture interac tive map Enable discrete access for navigating large scale pictures interactive maps Although discrete access is not that important for navigat ing picture interactive map it will also benefit if the infor mation space is really large and the user needs to locate a place far away from the current position 69 5 2 A comprehensive look at the navigation methods in digital media deur aAnoerojur aqnjord 103 aoeds UBISIQ e am I oe mmm A 70 5 Design space Notations 1 Mouse Stylus Touch screen Scrollbar 2 Mouse Stylus Touch screen Tap and drag 3 Mouse Stylus Touch screen Touch and go 4 Mouse Stylus Touch screen Planar 5 Keyboard arrow keys for moving and key for zoom in out 6 Joystick 7 TWEND 8 Search Box 5 2 A comprehensive look at the navigation methods in digital me
105. order to analyze the performance for navigating video of different length and to average the task completion time the length of the three pieces of video selected for the formal tests are 20 minutes 20 minutes and 50 minutes And the targets are at around 5 minutes 15 minutes and 25 minutes respectively Procedure For each of the two navigation methods the user is asked to do navigation tasks on the three selected pieces of video The order of the two navigation methods and the order of the three pieces of video for each navigation method will be pseudo randomized so as to counter balance the learning effect of the user Before the formal tests of each navigation method the user will have the chance to practice with the coming naviga tion method on the practice video After finishing the prac tice test the user can start to do the formal tests The user is asked to press F1 to load the video and start playing 7 2 Participants and time recording and to press F2 to confirm the target is found which stops the time recording as well The program will check the correctness of the found target by checking the current time Only if the target is correct is the user able to proceed to the next task Otherwise the user is asked to continue with the current task and the time recording will not stop Special requirements Confirmation of finding the target User will be required to confirm finding the targeted time point as soon as
106. ormed six dimensional design space can be gener ated the Access dimension is divided into two sub dimensions in the design space according to the two groups described before The classifications of the dimensions which describe the design space are consistent with the given taxonomy of navigation which classify the proper ties of navigation With the structure of the design space every kind of navigation property discussed in the previ 43 44 Overview of design space Shading implication 5 Design space ous chapter can be presented by a square As shown in Figure 5 1 the three dimensions in the lower part of the design space describe the Input Control and Manipulation properties the upper part describes the first group of the Access property and the right part describes the second group of the Access property the left part describes the Navigation Search dimen sion which represents the target of the navigation task With the six dimensions the design space is consisted with 96 squares which represent 96 combinations of navigation properties Out of the 96 combinations there are 24 com binations which are not possible They are represented by trellis shading in the design space and are called blind spot of navigation in the work More detailed design ra tionales will be discussed in the next section 5 1 2 Design rationale As can be observed from the design space the color deep ness of t
107. ots and represented by trellis shading Hence there are totally 72 valid squares in the design space 49 Design rationale of Navigation Search Blind Spot 50 Color Metaphor Effort indicator 5 Design space Color Metaphor To clearly distinguish the dimensions in the design space different colors are used to indicate dif ferent navigation properties The three dimensions at the lower part of the design space address the control the nav igation They are indicated by blue color with the color from light to dark to indicate the features along the naviga tion process The Access properties which are at the top and right side of the design space are indicated by orange color And the search types which are at the left side are indicated by purple color 5 1 3 Notations and representation Before presenting the notations and representation used in the design space another navigation property which has a big effect on the navigation performance is going to be introduced This property measures the amount of efforts a user needs to pay when executing a navigation task using a navigation method This property is classified into four levels as following Direct positioning Confirmation needed With the help of predefined frames Totally manual search From the first to the fourth measure the efforts the user needs to pay for finishing the navigation task increases Direct positioning means there is no manual sea
108. p donne PEPER umi map gm tum ka nem en m dus aug enmt quer Era er irs mh aem Dan my pen qm nq pasos io A y en a PA mm ER Y Re EE Hypo me DEE qnom mms Los LE Loses LEE n 105 7 3 User test program OSPIA L ISIP ur 1opi S eur ourn JOYSUIIMS WEIBOLA ISA 8 Z JINSIA AAA OA uua Mi AEZH var Ki 47145 bar epee tr a endur penads Suruyap sanba Agman lt JUAWAUYAJ BIT JO 99491 ft SSO WUDISISUOD SASLQWEP sog VONBINBYLON jo SINUEWAS y xeu S AJUaISISUOJ uondinsaq Amqeasipaad unb 01 pwy nuaw 1431400 204 yap u suon t pqx asnoui sues aq 03 pan SWONJE enua AJua sISYOJ LIJOIABYIG oye Surmuta sonp9y sdde ssou e uoneuuoju Eau ZuIpOD Joy ADUDISISUOD ENSIA suonu9sAuo Zeue Y MOPUI AA 2 anviwon sul SES EZ Dada dwo vir tef 91 ui puy nni antye Z 1 tTtwid og mai peo t t 119814 enia o pit HT Dar Agra ur AOA rent 204 punaj 84 oj papa si br Mag sous piuinbas 943 295 PGA se buo Ty Latz aput 343 bag au DW avi PUY 04 posu Lang NOA S LA MODUM JO ANUSA Vil INDgE Syn uS tt vu Day T WEE mouv 7 Validation 106 O9PIA ejiBrp ur OIJUOD paseq aje1 uojjnq PULMAL PIEMJOJ 1961 JoysuaaDs ure1801d ISI 6 Z BINS x Aa RR w20 ew PAPA 187 4 f j AP ji pum my AA MT BI me AZojouyzan ang st dno48 ano 3noqw snqej s ssej epo maiS DID ES se kaban an pug fow saye P4 SEd depi ayi ETO ai L 1 5304 ln o Lang pura PONA Lag BE
109. pT t98OEL 7 ALOGC PT DG krccn 8IPOL Tt 2 metre Jager eso rrocasee James bass Bros Jas best amer leecasys toast ra See erc jesaverser ra ect besetz ec best recover ereere E9 CU BEGDIZ EE CONTENER ezt1t6 GE LG EPOELTI ooron b ETAGEZZI OeevIE OT PSELESE ACELITE EDGTO 29 G O essersant berg ssscrez ac R R are aert ekiLEEST LRR G Een recsersor rzemsrer reng rest es R TR Feet CR CRS LO JEECELE ED LTH TUH LOTH EZILECLT C TH Iogcee EDEGLET GL CT LCH 24 ees ae R Jos sz ema less resorsr R emere esate VS G bs sees Pas rosa _jesrsewst jasseszes jescesser jarssaser R RRR TIGOTTO1 v9rci96 v EGEGLEGET geoes LT en riz 0i ELEGLZ EZ L AE LOPGZ GT LTH ergyzc THA GO jesseseer ba Fer Jeer Konz us Jeer assores R Foz Jeer Lo 279001 secostor eccarera recoerte Sg jesszeroe zereo jarsezaer comeron jatzzaoe Jero wasa Eti T97TGE TT 270 69 02 e amp czeros Iti amp 0 6E LOTH Iegergco TIPIT OS EGBECTE TOURED BT oct ELZEZ E FITELG S 24 C jaserestr Fees cceosast_ rrocsotr O aert Jesse H agesos _jesssere estos ti ee C ce oatucoet eet jeesmsor racoraet jezezeszt jeoinzem jeractear teenscc jezuore tocara szocrs osse stoivz TT EBOSOE GO Ire BGET LZ Jesper verbeter L 31 POr t BFTLLISE rg 2686125 22 G eostescs ees jserseste jerssesrst base jvorsrose jerenssrse_ ses jeseree rrsero orrors re CR O T rosserie ae beer coserct mzeree az Kurz est roce T EE WEE Se
110. permutations for the four navigation methods for digital document 24 versions of the test program were developed The permutation of the nav igation methods for interactive map and digital video just repeats for 4 times and 12 times respectively Figure 7 1 7 9 show the screenshots of the test program for every navigation method The benefits of running the user test by the test program include the following points Navigation method limitation The program only provides one navigation method for a task Therefore the user is restricted to use the required navigation method to perform the task and thus the re sults will not be impacted by using the undesired naviga tion methods Automatic target checking The correctness of the target found is checked by the pro gram The check for digital document is done by check ing the current page the check for interactive map is done by checking the latitude longitude value of the found tar get and the check for digital video is done by checking the current time If the target found is incorrect the program will give feedback to ask the user to continue and the time recording will not stop Only if the target found is correct will the continue button be shown in the program win dow and can the user proceed to the next task Precise time recording The time is recorded by the program instead of by using a stopwatch Hence the time recorded is more precise 7 3 User test program 9
111. put 24 Relative input 24 175 176 Index Input device 6 Intelligible audio feedback 19 Manipulation 26 Direct manipulation 26 Tool based manipulation 26 Navigation causality 53 Navigation control 7 Tool based navigation control 7 Navigation method 6 Navigation stage 80 Navigation state 80 Navigation Search 28 Arbitrary browsing 28 Targeted coordinate playback position search 28 Targeted object position search 29 Optimal path 85 Orientation cue 67 Reachable state 84 Required state 85 Semantic direct manipulation 62 Shading implication 44 SmartPlayer 17 Standard block diagram 85 Tap and drag 20 Time compression 18 TLslider 17 Touch n go 20 TWEND 11 Visualization of audio 72 Zoomable User Interface with Overviews 21 Typeset November 18 2009
112. rases keet emt rese rwr ere _ ssemevt eme rosa Is lervere te ees Ja jsencares keet az el PE H Fer R jonsnews catorree JS meotaret janzesne rtrcre et znerocor vtt LATER ster er Kee U reege ALH LSkPTYEH 66ctz9 cz BETEIZ OZ 9retot 6 gt ee nctz Iess E AT faresere rocrrer _ vecarcct_ revsver Jess sr rer Jas rose ees Ft escent eur Eeer Kaes ee Eeer eacewor ratmeet rovioree eccronct ER jssworc tort recemo vt ster SALIZT ET EILTOYT EFEETTOT UDC TE CL Ice ABLPOTE GI IPILAEZT enor troonQ zz Ise E M Est Jsortsczz earet rare Dr Scores ac gene ee seoenoYr Tea UR jezrosee meze Laser IER UGT tee IlELELTY IDBOESOOT rEsPOYE Lee est PZLEEZE TIPENP T IZDEEQ E ILODITO Iiceet OT beer zosssres zz CH G H H Jaar Esc H GH Eee vara vounn vonn EIE AE ee e E Raw test data 164 juoumoop EITAIP 10j e3ep 159 Mey CH SMS TODO TE Jeder jorz 6 ie FLEPLL Er It eier el EGDITEOT L ALIA et ZTLECTE Jose Jeer Ir IESSE ssoTtt ee meer SCOTT Or jestrer tr trae er streorsr rsrstrs jetzvoe sc rrecwo tt rer LH LS EH better ressst9 reer vrceeec Josee wtresrtr rsot ys Jerez aerer Jaeger zeroscs W PISOTZZ 221006 BOT art EOTEYG ES IZEZETLT JISCOsOTE GEES br BETSEZ GT Ire kee Joere Jeer EER 1SsEES S arr Freet seat er Imrooes sovesS ET risuawor Iesse estoy eewer Ier wcsrses rez Biest or Jecenccs tzecesow EXTST SET RLLTSS OT eran resseg wc oereos s Joss I
113. rch ef fort from the user s side is needed after telling the machine what he she wants to have In this case the navigation method will take over all the search responsibilities and find the unique correct target for the user For example if the user type a page number in the page number input box and press enter the navigation method will definitely lead you to the targeted page of the page number given No extra manual effort is needed 5 1 Design space Confirmation needed means the effort needed from the user s side is to confirm the correctness of the search result given by the navigation method or to select one from the several possible results the navigation method provides In this case the navigation method also does the search work for the user according to his her requirement but the user has to confirm whether the given result is the desired one or to choose one from several candidate results For exam ple if the user types some keywords in the search box the navigation method will lead him her to the position where the keywords are included But since there may be several positions have the keywords included the user has to con firm whether the first given one is correct If not he she has to choose the desired one from several candidate re sults With the help of predefined frames means the user has to search for the target manually often visually or acousti cally But the navigation method provides
114. rg ur L pan A iiin P w e ha 1 m si um gen AA pran y antes a pe A A 31205 WK WW ef MADE 157 OapiA ejiBrp 103 papeo Surids oou ammys jo ure18erp 323010 uorrenfeag 9 q am Bag N a N p 1 morenn D E Atty j Ni EES Uem T d gt AA andan F A an N e No 4 F l pss r Y mou ey D Evaluation block diagrams for digital video 158 OSPTA e1I2Tp 107 JaauM YAN Terp Sof jo urerSerp xpo q uoneNTEAg Z A JANB 159 OSPIA TejiBrp 107 UOSLIG Jo ure18erp xpo q Uonen eAq g a am er ee gen m u woche NHS A E La Appendix E Raw test data 161 E Raw test data 162 1 juoumoop eyiBrp 105 eyep 1991 Mey T A AMBI C H H rises ecortre poen R sees sc Jerseros rrsereze az oss D H Fees kreeg overere recoger jassserte zorooroc lacwscer kent sz stores eo Secret BEDEEG PZ etc AIDIGEA er ETGSOPT WE ZDLES ET IGDBEYGLZ LEOGOS EW LEETE tz EEGOGS E eviLDGL BIZISE 2 6 B R Fees sessie va L R laccecrat_ erstce tr eeeczoue East az Jeer E cvoncot are Kat Jeans sorterer R sa rermrr rsessr sorses en porre te siZOG ET Josee ie czirz GONMT OT resto ogv 9 LODETE ET IZOPITE SEILSYL FQDELEL 24 frrsseret_ ecercros risstoss jsssensem svcowse beer recsacse Jeer ees serrare ear rass Emer Er ess lorverare emere er stomstet ert Jess leresooce lavewee ac recetes ed ser ILLEZTE IZZAONTL eet IZOOTQ GT G POSSL LEEVI
115. rtPlayer can detect scene complexity and predefined events the playback speed is adjusted higher in the less important segments while play back speed can only remains slow when using the tradi tional player since the user do not have any information about the video contents Lauer and H rst 2007 proposed several designs to fa cilitate navigation for audio based educational multimedia contents some of which are explained below 3 2 Navigation method Time compression An additional slider was integrated in the player which enables user to adjust the playback speed between 0 5 times and 3 0 times of the original speed Be sides they integrated an algorithm to detect and remove the speech pauses that are longer than a certain threshold value Elastic audio slider In order to allow users to do fre quent playback rate changes the author integrated an elas tic audio slider in the progress bar Figure 3 10 It allows the acceleration or slowdown from the preset audio speed by dragging the slider to the right or left along a rubber band The speedup or slowdown factor is determined by the distance between the mouse pointer and the slider thumb which is called tension However it sacrifices the truly mapping between slider position and audio time H rst et al 2004 Audio progress 21 x ME ly 0 min 5min 10 min 15 min 20 min 0 6x 0 6x 1 5x lid BE ti Figure 3 10 Elastic audio slider
116. s and to evaluate the usefulness of a novel navigation design A user test run by a program is included in this work to support the framework proposed Because of the time and resource restriction the user test is only designed to vali date a part of the framework that is important More user studies could be done in the future Following is a brief summary of each chapter to provide an overall structure of the paper 1 Introduction 2 Background has a background look at the digital me dia navigation and design space It gives an overview of what the navigation in digital media is about and what a design space is 3 Related work lists other related publicized work which is about creating design space novel navigation and interaction or evaluation results and user study of naviga tion methods for digital media 4 Navigation of digital media gives a taxonomy of nav igation explores the current commonly used input devices and navigation controls and presents all the navigation methods which will be analyzed in this work 5 Design space presents the design space describes the design rationale and explains the notations and analysis methodologies A comprehensive look at the current nav igation methods and future design suggestions are stated here 6 Evaluation block diagram presents the generation and evaluation methodologies of using evaluation block di agram 7 Validation describes the user
117. s the human factors of computer graphics interaction techniques In IEEE Com puter Graphics and Applications volume 4 pages 13 48 IEEE Computer Society Press Los Alamitos CA USA November 1984 Dan B Goldman Chris Gonterman Brian Curless David Salesin and Steven M Seitz Video object annotation navigation and composition In Proceedings of the 21st annual ACM symposium on User interface software and tech nology pages 3 12 Monterey CA USA 2008 ACM Gero Herkenrath Thorsten Karrer and Jan Borchers Twend twisting and bending as new interaction gesture in mobile devices In CHI 08 extended abstracts on Human factors in computing systems pages 3819 3824 Florence Italy 2008 ACM Ken Hinckley Edward Cutrell Steve Bathiche and Tim Muss Quantitative analysis of scrolling techniques In Proceedings of the SIGCHI conference on Human factors in computing systems Changing our world changing ourselves pages 65 72 Minneapolis Minnesota USA 2002 ACM Wolfgang H rst Tobias Lauer and Georg G tz An elastic audio slider for interactive speech skimming In Proceed ings of the third Nordic conference on Human computer inter action pages 277 280 Tampere Finland 2004 ACM Takeo Igarashi Speed dependent automatic zooming for browsing large documents In Proceedings of the 13th an nual ACM symposium on User interface software and technol Bibliography ogy pages 139 148 San Diego California United St
118. same navigation method are con nected by black lines If the circles belong to the same navigation method they are connected by black lines The connected circles and black lines are regarded as a whole to represent a naviga tion method 5 1 Design space Dashed lines with arrow represent navigation causality If the work of one navigation method will trigger the work of another one then a dashed line with an arrow drawn from the first navigation method to the second one with the arrow pointing to the latter one is used to represent this navigation causality Since the circles connected by black lines are regarded as a whole to represent a naviga tion method the dashed line can be drawn from any circle of the first navigation method to any circle of the second one For example when the user uses tap and drag to navi gate the document or map the thumbs in the scrollbar will also move correspondingly Thus there exists causality be tween the two navigation methods Drag and drop can cause the work of scrollbar And therefore a dashed line is drawn from a circle of tap and drag to a circle of scroll bar Figure 5 2 If the second navigation method is also the cause of the first one then a dashed line with double arrows should be drawn between them An example could be scrollbar and page number input box 5 1 4 Analysis methodology and guidelines When analyzing the navigation methods designed for an application th
119. sign ods for digital video nowadays Therefore a main task for suggestions for designing new navigation methods is to design and im navigation in digital prove direct manipulation techniques Direct manipulation video is especially useful and important for accurate video navi gation for professional video processing like analysis cut ting and annotation etc Therefore exploring the design in this area is worthy to be done by designers Besides mak ing the current navigation methods more efficient could also be tried Some suggestions are proposed below Improve direct manipulation technique Although there have been several direct manipulation tech niques developed they are still not so efficient and flex ible Some require preprocessing which do not support real time interaction some can only manipulate the prede fined objects Therefore improving the techniques to make direct manipulation as flexible and easily to be integrated as tool based navigation methods is essential 76 5 Design space Implement random access and predefined frames into di rect manipulation Today s direct manipulation techniques are still simple se quential access by dragging the objects in scene Hence it is suggested that the feature of random access can also be introduced to direct manipulation And the system can also provide some predefined frames which can be accessed by direct manipulation techniques Design dynamic pred
120. sign suggestions The biggest problem for navigating a large scale and high information density picture or an interactive map is that the user has no information about the position distance and direction of the desired target in most cases Thus search ing for the target in such a complex environment will be exhausted and always depends on luck Therefore pro viding helpful information or cues is very important when implementing navigation methods Some proposed sug gestions are given below Use color metaphor or shape metaphor etc to classify ar eas Color metaphor or shape metaphor can limit the areas user needs to check thus to reduce user s visual effort and to make the search faster For example Google map uses red color to indicate hospitals and green color to indicate parks etc 5 2 A comprehensive look at the navigation methods in digital media Highlight important information or provide quick access to the important targets When having a search task the user often has some infor mation about some nearby places e g the target is near the central railway station or the place can be reached by taking a certain bus to a given stop etc Hence highlight ing some important information or providing quick access to some important targets may give user semantic informa tion about the possible target position if the highlighted targets are related to the information the user has Dynamically indicate related infor
121. squares And they will be presented by different kinds of lines in the evaluation block diagram Both will be introduced in the following sections Notations and representation Figure 5 2 5 3 5 4 and 5 5 are the examples of the design spaces for the four types of digital media which include some of the navigation methods The notations and rep resentation in design space include the following rules Properties of navigation methods are circles in the corre sponding squares It has been discussed in 5 1 1 Overview that each square totally 72 valid squares in the design space repre sents a combination of some navigation properties There fore when representing a navigation method in the design space a circle in a square represents that the navigation method has the properties indicated by this square To com pletely represent all the properties of a navigation method the designer can position circles in all the squares which indicate the properties of the navigation method Extra signs in the circles are the effort indicators To present the four user s effort levels which were just discussed extra signs are added in the circle No sign is added if it represents direct positioning a plus sign is added if it represents confirmation needed a smaller cir cle is added if it represents with the help of predefined frames a cross sign is added if it represents totally manual search Circles belong to the
122. t and rooms for improvement To address these problems this thesis work proposes a systematic way to evaluate the navigation methods in abstraction for digital documents picture interactive map digital audio and dig ital video The work gives a taxonomy which explores the properties of navigation in digital media based on which a design space is created Then the work has a comprehensive look at the major existing navigation methods for the four types of digital media by using the design space analyzes the current design problems and proposes some future design suggestions Afterwards an evaluation tool is proposed to measure the capabilities of the navigation methods for required navi gation tasks With the tool designers are able to evaluate the suitability of a navi gation method for a task and to compare different navigation methods A user test is included in the thesis work which is to verify the correctness of the design ra tionale of the framework and the results derived from the evaluation tool It gives the confidence that the design space and evaluation framework is useful for the designers Xx Abstract xxi berblick Durch die Entwicklung der Computer Technologie spielen die digitalen Medien wie z B elektronische B cher interaktive Landkarten oder digitale Musik und Video eine immer bedeutender werdendere Rolle in unserem heutigen Leben Da die digitalen Medien sehr schnell und einfach bereitgestellt getauscht
123. ta tion of digital video fluid interaction techniques for the control and annotation of digital video In Proceedings of the 16th annual ACM symposium on User interface software and technology pages 105 114 Vancouver Canada 2003 ACM Liyang Sun and Francois Guimbretiere Flipper a new method of digital document navigation In CHI 05 ex tended abstracts on Human factors in computing systems pages 2001 2004 Portland OR USA 2005 ACM 173 174 Bibliography Shumin Zhai Barton A Smith and Ted Selker Improv ing browsing performance A study of four input devices for scrolling and pointing tasks In Proceedings of the IFIP TC13 Interantional Conference on Human Computer Interac tion pages 286 293 London UK 1997 Chapman amp Hall Ltd Index Access 27 Continuous access 27 Discrete access 27 Random access 27 Sequential access 27 Analysis guideline 58 Blind spot 49 Card s design space 9 Color metaphor 50 Connecting line 82 Consequent leading line 84 Control 26 Position based control 26 Rate based control 26 Design space 8 43 Digital media 5 DimP 21 Dragon 15 Effort indicator 52 60 Effort level 50 Confirmation needed 51 Direct positioning 50 Predefined frame 51 Totally manual search 51 Elastic audio slider 19 Electronic media 5 Evaluation block diagram 8 79 Evaluation preference 86 Footprints scrollbar 11 Halo 15 Input 24 Absolute in
124. te 5 Mouse Stylus Touch screen Fast forward Rewind button frame skipping 6 Shuttle wheel Spring loaded 7 Jog dial Click wheel 8 Dragon Chapter 6 Evaluation block diagram So far the work has analyzed the navigation in digital me dia introduced the design space and had a comprehensive look at the current navigation designs for the four types of digital media In this chapter another useful tool in the framework evaluation block diagram is going to be ex plained Evaluation block diagram is a kind of navigation flow dia gram which is used by adding connections between prede fined states to examine the navigation capabilities by paths formed in the diagram The states in the evaluation block diagram are the navigation properties in the design space and having a certain property is represented by having a connection to the corresponding state The purpose of introducing the evaluation block diagram is that instead of examining the navigation capabilities by checking the squares in the design space it can be more easily and intu itively observed in the block diagram by looking at which states are reachable and what the possible paths are Given the information of required states or optimal path it can be easily used to measure whether a navigation method is able to accomplish a navigation task And it is very easy to compare the different navigation methods and measure the suitability by comparing with the gi
125. te position search not possible 5 Keyboard arrow keys for moving and key for zoom in out Relative input Position based control Tool based naviga tion Continuous amp Sequential access 6 Joystick Relative input Rate based control Tool based navigation Continuous amp Sequential access 7 TWEND Relative input Rate based control Tool based navigation Continuous amp Sequential access 42 Navigation methods 8 Search Box Relative input Position based control Tool based navi gation Discrete amp Random access only targeted object position search possible 35 4 Navigation of digital media 36 deur aAnoerayur aqnjord 107 spoujour uoneSraeu jo Kreurumg Cp 91314 42 Navigation methods 4 23 Digital audio 1 Mouse Stylus Touch Screen Timeline slider Absolute input Position based control Tool based navi gation Discrete amp Random access Enhanced Mouse Stylus Touch Screen Zoomable timeline slider Mouse Stylus Touch Screen Position based navigation with intelligible audio feedback 2 Mouse Stylus Touch Screen Fast forward Rewind button increasing playback rate Relative input Rate based control Tool based navigation Continuous amp Sequential access 3 Mouse Stylus Touch Screen Fast forward Rewind button frame skipping Relative input Rate based control Tool based navigation Discrete amp Sequential access targeted playback positio
126. tent to search for the desired target rate based con trol can increase the navigation speed in less interested places And if the user has to sequentially navigate to a tar get far away from the current position rate based control can reach the target at a higher speed than position based control However it may be not true if the user needs to navigate a very short document or a very short video clip In this case speed does not have advantage any more In stead position based control will be easier to control than rate based control And it is also more accurate than rate based control if the user needs to locate precisely to a spe 5 1 Design space cific position Nevertheless rate based control is still laid at the right side of position based control since the navigation length is not that short in most cases and the length of the document audio or video nowadays becomes longer and longer Thus we believe that in most cases rate based con trol can outperform the position based control in sequential access There are some empirical data to support this argu ment The experiment results in Hinckley et al 2002 presented a crossover effect to prove that when the document length becomes longer the rate based scrolling will outperform the position based scrolling The user study in Zhai et al 1997 showed that rate based scrolling is significantly faster than position based scrolling for navigating web based documents
127. teraction it can better support the tasks such as video cutting scientific footage analysis detailed sport scene review and video annotation etc sone Figure 3 7 DRAGON 3 2 Navigation method Ramos and Balakrishnan 2003 designed a TLslider Twist Lens slider a variant of fish eye slider for use with pressure sensitive digitizer tablets to navigate video As il lustrated in Figure 3 8 the TLslider is a sequence of thumb nails which provides a visualization of the complete video stream The thumbnails are mapped to the corresponding video frames To overcome the thumbnail occlusion prob lem in regular fish eye sliders the TLslider morphs the lin ear layout to an s shape layout and the pen s pressure de cides the sinusoidal amplitude of the lens Twist Lens Pressure Edo 0 e ug ATT e lt a L E Figure 3 8 TLslider Cheng et al 2009 designed SmartPlayer Figure 3 9 an adaptive fast forwarding interaction model adopted by the metaphor of scenic car driving which helps people to quickly browse videos with predefined semantic rules It adjusts the playback speed based on the complexity of the scene and predefined events The player can also learn user preferences about the predefined event types and adjust adaptively to the suitable playback speed The most two novel parts in the design are the scented seeker bar and predefined event list The visual scent on the seeker bar is encoded by t
128. test which is designed to evaluate the validity of the framework followed by the results and conclusions 8 Summary and future work summarizes the work done in this thesis work discusses the findings and the work that could be done in the future Chapter 2 Background 2 1 Digital media The digital media we are talking about and interested in this work is those media types that are opposed to tradi tional media types which are often based on some phys ical mediums like paper Digital media is also usually re ferred as electronic media That implies the mediums for carrying digital media contents are usually electronic de vices like computer mobile phone PDA etc Nowadays the most commonly used types of digital media are digital documents pictures interactive maps digital audio and digital video etc As mentioned earlier comparing to tradi tional media digital media has the advantage that it can be easily distributed shared carried accessed and archived A large amount of digital media contents can be obtained for free or at a low rate Hence people tend to use digital media contents more and more nowadays it is even replac ing the traditional media And therefore digital media has become an important and increasingly interesting research topic these days Concept and examples of digital media Navigation method 2 Background 2 2 Navigation designs The navigation tasks in digital media are usually
129. the kinds of navigation which are not able to be done by the navigation method will be mentioned 29 30 4 Navigation of digital media In addition some enhanced techniques will also be listed below the original navigation methods Although these en hanced navigation methods will not be directly included in the design space since they are in the same category with the original one some of them will be used for fur ther analysis After listing the navigation methods for each kind of digital media a figure summarizes the properties of the navigation methods for this kind of digital media will be presented 4 2 1 Digital document 1 Mouse Stylus Touch screen Scrollbar Absolute input Position based control Tool based navi gation Continuous amp Discrete amp Random access Enhanced Mouse Stylus Touch screen Semantic scrolling Mouse Stylus Touch screen Speed dependent auto matic zooming Mouse Stylus Touch screen Rapid serial visual presen tation 2 Mouse Stylus Touch screen Thumbnail enhanced Scrollbar Absolute input Position based control Tool based navi gation Discrete amp Random access 3 Mouse Stylus Touch screen Thumbnails Absolute input Position based control Tool based navi gation Discrete amp Random access Enhanced Mouse Stylus Touch screen Space filling thumbnails 42 Navigation methods Mouse Stylus Touch screen Bookmarks 4 Mouse Stylus To
130. to their physical properties and the number of spatial dimensions In Mackinlay 1986 the author applied the morphological design space analysis technique to human computer inter face design He developed an application independent pre sentation tool that generates graphic presentation designs 3 2 Navigation method 3 2 Navigation method There are a variety of research projects proposing ideas for improving the existing navigation methods as well as some others creating new mechanisms for navigating dig ital media By reviewing the existing designs the clas sification in this work will be more complete Some re searchers also did user studies to compare their improved or novel navigation methods with the currently commonly used ones which helps the development of the analysis methodology in this framework with empirical results The following section will present several improved or novel navigation designs which are interesting and may have big effects on the development of digital media navigation Alexander et al 2009 designed a Footprints scrollbar an enhanced scrollbar to help people return to the previ ously visited document regions As shown in Figure 3 2 the Footprints scrollbar include colored marks to indicate the previously visited areas with cold colors indicate an increasing age than the hot ones A small thumbnail will be shown when the mouse is over a mark and a big ger thumbnail will be shown when th
131. uch screen Tap and drag Relative input Position based control Direct manipula tion Continuous amp Sequential access 5 Mouse wheel position based Jog dial Click wheel Scrolling Relative input Position based control Tool based naviga tion Continuous amp Sequential access Enhanced Mouse wheel position based Jog dial Click wheel Se mantic scrolling Mouse wheel position based Jog dial Click wheel Speed dependent automatic zooming 6 Mouse wheel rate based Scrolling Relative input Rate based control Tool based navigation Continuous amp Sequential access Enhanced Mouse wheel rate based Scrolling Semantic scrolling Mouse wheel rate based Scrolling Speed dependent au tomatic zooming Mouse wheel rate based Scrolling Rapid serial visual pre sentation 7 Keyboard up down button Scrolling Relative input Position based control Tool based naviga tion Continuous amp Sequential access Enhanced Keyboard up down button Semantic scrolling 31 32 4 Navigation of digital media 8 Keyboard arrow keys Thumbnails Relative input Position based control Tool based naviga tion Discrete amp Sequential access Enhanced Keyboard arrow keys Space filling thumbnails Keyboard arrow keys Bookmarks 9 Keyboard page up page down button Relative input Position based control Tool based naviga tion Discrete amp Sequential access 10 TWEND Relative i
132. und gespe ichert werden k nnen verdr ngen sie nach und nach die traditionellen Medien Sie werden zu den bedeutendsten Medien f r Menschen um Informationen zu erlangen Zur gleichen Zeit besch ftigen sich Wissenschaftler damit die digi talen Medien einfacher und effizienter zu lenken Jedoch k nnten sie sich ver loren f hlen in der F lle der Gestaltungsm glichkeiten da es keine M glichkeit gibt die Lenkungsmethoden systematisch zu erproben Und die Entwickler haben kein klares Bild ber die derzeitigen Verbesserungsm glichkeiten Um diesen Problemen zu begegnen entwickelt diese Abhandlung einen systematischen Weg die Lenkungsm glichkeiten zu erproben f r digitale Dokumente Bilder und in teraktive Landkarten sowie Musik und Video Diese Abhandlung gibt eine Systematik der Eigenschaften der Lenkungsm glichkeiten der digitalen Medien basieren auf dem jeweiligen Design Die wichtigsten existierenden Lenkungsmeth oden werden beschrieben f r die vier Bereiche der digitalen Medien durch die Nutzung des Designs der Analyse der derzeitigen Designprobleme und durch das Vorschlagen zuk nftiger Designs Zum Schluss wird ein Evaluierungstool vorgestellt um die M glichkeiten der Lenkungsmethoden f r die gew nschten Aufgaben zu messen Mit diesem Toll haben Entwickler die M glichkeit die Brauchbarkeit der Lenkungsmethode f r eine Aufgabe zu bestimmen und ver schiedene Methoden zu vergleichen Au erdem enth lt die Abhandlung einen Be
133. users to have a concept of what the page is about without having to navigate to this page first 5 Design space juoumoop ej18rp 103 adeds UBISIq 9 6 JINSIA EES SEE SE SEE M i B EE E m 5 2 A comprehensive look at the navigation methods in digital media 65 Notations 1 Mouse Stylus Touch screen Scrollbar 2 Mouse Stylus Touch screen Thumbnail enhanced Scrollbar 3 Mouse Stylus Touch screen Thumbnails 4 Mouse Stylus Touch screen Tap and drag 5 Mouse wheel position based Jog dial Click wheel Scrolling 6 Mouse wheel rate based Scrolling 7 Keyboard up down button Scrolling 8 Keyboard arrow keys Thumbnails 9 Keyboard page up page down button 10 TWEND 11 Search Box Find function 12 Page number Input Box 13 Page number increase decrease button 66 Current design situation of navigation in picture interactive map Future design suggestions for navigation in picture interactive map 5 Design space 5 2 2 Picture Interactive map Current design overview By examining the design space for picture interactive map Figure 5 7 findings obtained from observing the current navigation method designs include Most navigation methods are tool based few are with di rect manipulation Most navigation methods are continuous access few are discrete access Almost all targeted object position search methods require user s manual visual search effort Future de
134. ven optimal path What is evaluation block diagram 79 80 Structure of evaluation block diagram Three stages in the evaluation block diagram 6 Evaluation block diagram 6 1 Structure The structure of the evaluation block diagram is shown in Figure 6 1 The block diagram is generated from the design space The circles in the block diagram are called states which correspond to the properties in the design space Therefore the evaluation block diagram and the de sign space are fully consistent with regard to the structure and the evaluation results as well The states in the evaluation block diagram which are the properties of navigation methods are arranged from left to right according to the navigation process with the right most state represents the final purpose of navigation Therefore the states are divided into four stages along the navigation process which are indicated by four dash line frames in Figure 6 1 The first stage is the start stage of the navigation which is the input and manipulation One of the first two circles and one of the second two circles are connected to represent the combination of the input and manipulation properties There is no consequential relationship between the first two groups of states The second stage indicates the access properties of the nav igation As discussed in 5 1 2 Design rationale sequen tial access is the subset of random access since any nav igati
135. wind button is 242 677879s which takes a lot time the average completion time of using time line slider is only 25 032224s User test results Doc 1 s Doc2 s Doc3 s Average s Mouse Scrollbar 10 798084 35 930096 57 246605 34 658262 up down button 7 094392 27 131941 52 748366 28 991567 Mouse Thumbnails 6 823859 24 277129 45 386862 25 495950 Search Box 14 744578 16 510329 14 671899 15 308935 Table 7 1 User test results for digital document 107 108 7 Validation Map 1 s Map 2 s Map 3 s Average s Tap and drag 181 197867 282 636374 205 265415 223 033219 Tap and drag with cue 49 185602 110 415503 127 694658 95 765255 Search Box 20 195422 28 274294 20 702556 23 057424 Table 7 2 User test results for interactive map Video 1 s Video 2 s Video 3 s Average s FFW RWD rate control 247 523647 389 070588 91 439403 242 677879 Timeline Slider 30 513441 18 851420 25 683491 25 032224 Table 7 3 User test results for digital video Analysis of the user test results for digital document 7 5 Analysis and conclusions 7 5 1 Digital document As shown in Figure 7 10 for document 2 and 3 the middle length and long length documents the task completion time decreases along the four navigation methods For doc ument 1 which is the short length docum

Download Pdf Manuals

Related Search

Related Contents

取扱説明書  baixar arquivo - Brinquedos Bandeirante  User Manual - MCS Systems  VPCF13GGX/B - Clearance Club  Sony CMTDH3 User's Manual  Sony S2 User's Manual  

Copyright © All rights reserved. Failed to retrieve file