Magpie: User's Guide
Contents
1. The button is depicted as A in Figure 1 and if you click on it an initial Magpie toolbar appears showing a single button labelled Magpie and a little arrow pointer B Author Martin Dzbor KMi Page 4 of 14 E about blank Microsoft Internet Explorer provided by The Open University 6 0 spl j z Ioj x File Edit View Favorites Tools Help Back amp A isearch GejFavorites FPmedia G4 Ghr 4 Bi IE EF amp Address E about blank p Magpie Load Magpie Lexicon From local file Unload Current Lexicon From URL A Launch CS A4KTive Space Preferences B User identity Debug dump About Figure 1 Magpie toolbar as an integral part of Internet Explorer showing a button for showing hiding Magpie toolbar A and initial appearance of the toolbar with menu B Preparation ontology definition download Once you can get as far as described in the previous section Magpie 1s likely to be installed and registered correctly To test its functionality you will need to load at least one ontological lexicon A lexicon 1s a mapping from strings to entities within an ontology Domingue Dzbor et al 2003 Each entity in the ontology is referred by its instance name and class Instances may have more than one lexical representation For instance assume there is an instance known as BKD P in the ontology AKT KNOWLEDGE BASE Possible textual phrases that are to be associated with the2. a p i D al 1908 653506 08 653169 je modelling techniques to the web to r services For instance in the Lyceum S Project Details xpert users to jel Project Research Areas l ecti n g an d l Project Publications JOI Resulting Technologies ro ect we are i b Project Members to Shares Research Areas With ro ects j n A by of the ideas embedded in an academic technology to develop fhe web z a fae Local intranet F Clear Figure 9 Magpie collectors as an example of trigger services other projects relevant to the people mentioned on a page are listed r Author Martin Dzbor KMi Page 12 of 14 Semantic browsing history One of the purposes of the aforementioned data store and semantic log keeping is to manage users browsing history better and easier Since Magpie associates concepts and phrases found in the text with the actual URLs it enables us to browse the history using semantic means In contrast to the linear bookmark and history management provided by the standards web browsers semantic history is more powerful In addition to seeing the browsing history in a temporal manner our Where am I management tool is able to provide several complementary views e Organization by ontological concepts and phrases e Organization using iconic representation of web pages rather than URLs e Search of large history bookmark repositories using high level query language that focuses on web page
3. see pointer B in Figure 3 Similarly you can repeat the same clicking action on each of Magpie buttons Each button would highlight instance from the respective category in the web page The fully annotated web page is shown in Figure 4 and shows KMi projects research topics and areas of KMi and key organizational bodies More specifically Magpie correctly recognized phrase KMi as an organization see pointer A in Figure 4 and ClimatePrediction Net as Martin s project see pointer B in Figure 4 In addition several research topics areas are highlighted Ontologies Knowledge Management Knowledge Modelling and Knowledge Technologies all with light blue background in Figure 4 The fact that a particular category is actually highlighted is shown on Magpie toolbar by a relief button see compressed buttons pointed by C in Figure 4 E Members Microsoft Internet Explorer provided by The Open University 6 0 spi Eie x File Edit View Favorites Tools Help El q Back amp A GQsSearch Favorites Meda C4 A 4 AW S BH 4 Links A Magpie Person Project Research Area Organization p KNOWLEDGE MEDIA John Domingue INSTITUTE A John s Home Page ABOUT US NEWS PEOPLE 44 0 1908 655014 Contact John Domingue j g e f Members Pe 4 Affiliates si mnei Keys KMi Deputy Director Semantic Web Semantic Web S
4. Dzbor KMi Page 9 of 14 The next step is choosing one of the options in the displayed semantic menu Let us start with the simple one first Which other research areas are related to Knowledge Management Translated and abbreviated into semantic services labels this question corresponds to the first option which is shown by pointer A in Figure 6 In response to requesting this particular semantic information about phrase Knowledge Management a small window pops up resembling the one in Figure 7 According to AKT ontology areas such as knowledge acquisition knowledge maintenance and so one are relevant in the context of clicking on Knowledge Management 23 http plainmoor open ac uk 3000 request op service name k OW LE Oj x Bak gt G A A Asearch Gyravorites SMedia z Links B A Magpie The research area Knowledge Management is related to the following Knowledge Acquisition Knowledge Lifecycle Knowledge Maintenance Knowledge Modelling Knowledge Publishing Knowledge Retrieval Knowledge Reuse Figure 7 Response from an ontological reasoning engine to the query about other research areas related to Knowledge Management E http plainmoor open ac uk 3000 request op service name d Back amp e A Qsearch Favorites Efmedia T Links i A Magpie John Domingue shares interests with Dnyanesh Rajpathak Knowledge Reuse Problem Solving Methods Ont
5. KNOWLEDGE MEDIA K INSTITUTE Magpie User s Manual and Functionality Overview by Martin Dzbor http kmi open ac uk people dzbor 2003 2004 Knowledge Media Institute Executive summary This document describes a tool that assists users with browsing the Web Browsing activity involves finding the right web page and making sense of its content Most effort so far was focused on finding web pages using standard information retrieval mechanisms Less attention was given to finding a web page visited in the past browsing history and even less to the sense making process Magpie as a suite of tools provides support for the interpretation of web pages and semantically enriched browsing history management It automatically associates an ontology based semantic layer to web resources allowing relevant services to be invoked within a standard web browser Relevant research papers are Domingue Dzbor et al 2003 Dzbor Domingue et al 2003 More information about Magpie is available from the project s home page What is Magpie Magpie is a tool that plugs into a standard web browser Microsoft Internet Explorer Its novelty comes from the fact that through Magpie the web user can access not only the information provided by web page authors but also a significant amount of background information and knowledge Typically it is this background information the users are seeking on the web and sometimes it takes
6. aforementioned instance include Bayesian knowledge discoverer Bayesian knowledge discovery and BKD Similarly instance JOE DOE would translate to Joe Doe J Doe Doe J Doe Joe etc Lexicon containing the instances such as those mentioned earlier represents items the Magpie extension would be able to work with and find in the web page Moreover thanks to the lexicon and underlying ontology Magpie is able to offer a suite of semantic services that are dependent on a class of a particular instance For example our BKD P is known as a RESEARCH PROJECT specifically and PROJECT more generally Similarly TJOE DOE is classified as a SENIOR RESEARCH FELLOW IN ACADEMIA or more generally as a PERSON Hence semantic services related to the PERSON class would show when clicked on Joe Doe in the web page Lexicon comes presently 1 e in this beta release in form of two files The main lexicon has extension onto and contains the actual instances from the ontology The second file with the extension menu is self explanatory it contains the definitions and labelling of the semantic menus associated with the classes from this particular ontological lexicon Unless you want to create your own ontology you do not need to worry too much about the syntax of each file In the future releases we shall move towards standard web language s
7. al information about them we are currently working on several other tools These additional tools shall extend the foundation of current Magpie toolbar and provide more value added services to the user We are mentioning them here in order to see Magpie in a particular context rather than an isolated product The additional tools include e Mechanism for trigger or push services and e Mechanism for semantic management of browsing history and bookmarks Trigger services in Magpie All services mentioned so far in the previous sections are essential user driven The user has to do some action choose a particular concept or phrase and request what additional information s he would like to obtain This request then goes to the ontology based reasoning engine which answers the requested query In addition to the user requested services we also develop so called trigger or push services These behave differently from the requested ones The user subscribes to a particular collector or adviser service and Magpie sends whatever was found on each web page to a central data store Different agents monitor this central store and if a set of data occurs that complies with a particular pattern respective service is triggered In other words using data driven reasoning new knowledge is inferred This chunk of knowledge is then pushed to the user s desktop where it appears in a dedicated window Consider example in Figure 9 Using standard user
8. al lexicons stored e g in XML RDF or OWL formats Author Martin Dzbor KMi Page 3 of 14 Installation and basic setup We start with the installation and setup of Magpie extension on the user s computer The current configuration of beta release of Magpie assumes that working folder is C Magpie While the user may change the settings this folder will be used for the storage of web page snapshots as well as ontology storage The entire functionality of Magpie is implemented as dynamically loaded library DLL This library has to be registered with your operating system before anything can be tested In the case you are installing Magpie from an executable installer magpie setup exe the registration should happen automatically However the DLL registration may fail in the case you have a running instance of Internet Explorer If this happens follow the instructions below to find out whether a running instance of Internet Explorer is a problem Once there are no applications or processes of Internet Explorer running go to the directory you installed Magpie to e g C Magpie and run double click on the file register magpie dll bat Alternatively you may do the registration manually The following command must be executed in the DOS prompt launch the DOS prompt window using Start gt Run and type cmd a DOS window with a prompt should show regsvr32 C Magpie Magpie dll If the DLL registration was accompli
9. content rather than label URL or access time Figure 10 shows an example of the Where am I tool in practice This particular snapshot works with the same ontological lexicon as we mentioned earlier AKT onto Therefore we are able to make sense of the entities found in the web pages we browsed in the previous sections We organize history browser around the same categories of entities Person Project etc Similarly the tool accepts the entities from the given ontology to form a part of the user s query Thus instead of going blindly through an extensive browsing history our tool enables the user to ask directly what s he wants needs Assume the user wants to see a page which s he accessed last week and which was presenting some visualization work by some academic research institution however s he can t remember the names of the people institution or project It would be futile to ask all the possible combinations of the keywords occurring in the previous sentence However using our ontology driven history management tool the same can be done fairly easily Find me a page that satisfies the following conditions in a conjunction e academic X e has research interest X visualization e and X isa concept known in the ontological lexicon AKT onto Magpie Where Am URL Operations Ontology Operations 0040m and academic x has research interest x visualization isited Pages imag
10. ervices Organisational Learning Knowledge PROJECTS Modelling Software Visualisation PUBLICATIONS I m KMi s Deputy Director and I m interested in how semantic web and SOFTWARE can support the creation and sharing of knowledge In particular at present I am interested in how the ontological enrichment of web pages can support interpretation I am also interested in how semantic web services can be easily created configured and deployed Martin Dzbor Martin s Home Page 44 0 1908 858524 Contact Martin Dzbor B Keys GiMStePredictiomiNet Jabber Design theory Knowledge modelling Reflection and Framing Rao WiadGalManaeamant Semantic Web Figure 4 Fully annotated web page showing all Magpie buttons categories turned on and highlighted in appropriate colours people light green projects chocolate brown research areas sky blue and finally organizations light lilac Contextual semantic menus The next after successfully highlighting concepts and phrases in the text we want to invoke some of the semantic services The adjective semantic means that the actual suite of services is chosen on the class membership of a clicked on phrase Please Author Martin Dzbor KMi Page amp of 14 note that only phrases that were highlighted by Magpie would respond with a menu of semantic services To invoke a semantic menu you need to move with the mouse pointer over a particular entity and click wit
11. es Person Project Kmi Planet Planetonto _ Project Details Project Research Areas Project Publications i Resulting Technologies http kmi open ac uk projects Project Members E B Shares Research Areas With My Planet Web Page Neighbour Pages ResearchArea J Organization 7 7 a F mA 3 JE SERTE ga F ny f L X i http KMI C uK people motta koir TEYE 7 FEENEY e Media Institute At T he Open University Figure 10 Where am I bookmark and browsing history management tool Author Martin Dzbor KMi Page 13 of 14 As we see in Figure 10 in response to the ambiguous query from above our tool produced a small list of web pages pointer A shows their iconic representation Each of the web pages is annotated using the concepts from the used ontological lexicon AKT onto These concepts constitute what we call semantic footprint of a web page in the space of a given ontology and in Figure 10 they are in four rectangles B In addition each of the concepts behaves similarly as it would behave on a web page In other words it is possible to invoke semantic services for the concepts C and investigate the details Thus we can arrive at the information we were originally seeking without knowing exactly what it was and where it was At the same time the semantic awareness of our tool reduces the total number of relevant pages and their physically linked neighbourhood D References Domingue J M Dzbor e
12. h the right button For AKT onto four distinct semantic menus could appear depending on the category membership of a particular entity They are shown in Figure 5 Project Details Project Research 4reas Project Publications Person Details Related Research Areas Works with Resulting Technologies Relevant Publications Shares Interests With Project Members Related Stories Competitors Current Projects Shares Research Areas with Related Technologies Same Problem Or Product My Planet My Planet Related People Invite To Web Page A Web Page Related Projects B C Might Fund D Figure 5 Four types of semantic menus defined in AKT onto lexicon A for the category Person B for the category Project C for the category Research Area and D for the category Organization Now click on one entity phrase in the text denoting a person For example let us assume we right click on highlighted phrase Knowledge Management Since this phrase became visible after pressing button Research Area we expect menu with semantic services specific for this category option C in Figure 5 The actual appearance of how your browser may look like is shown in Figure 6 As you see the menu appears exactly where the user clicked with the mouse It has a usual appearance of contextual menus under given operating system E Members Microsoft Internet Explorer provided by The Open U
13. ies of a particular class found in the web page Currently all research areas that are known within the loaded ontological lexicon would be highlighted in sky blue While this 1s a good colour for a page with white background it may be inappropriate for say light blue background In such a case you may easily change the colour of highlighters To perform the change access the pull down menu for a particular button category and choose second option Change Highlight Colour as shown in Figure 2 A colour picker dialog would show allowing you to choose your preferred shade Author Martin Dzbor KMi Page 6 of 14 Using Magpie while browsing In order to get in grips with Magpie functionality we suggest starting with the organizational memory lexicon Org Intranet Knowl Management onto file This file provides a lexicon for the KMi reference ontology which describes the life of a research institute in academia and would help you identify various people projects research areas and their relationships as they are known in this particular institute To see Magpie with KMi lexicon in its native environment we suggest the following http kmi open ac uk people Type this URL into your Magpie enabled browser and wait until it appears on screen Once the site is visible you may load the KMi lexicon if you haven t done so yet See previous section for how to go about it There are essentially two basic ways for a user to interac
14. long time to find appropriate and relevant information Dzbor Domingue et al 2003 While there is a lot of information on the web it becomes increasingly difficult to o Make sense of large amount of information o Formulate the search engine queries so that they find reliable and relevant information and o Filter out reliable and useful information from the vast amount available The issue of sense making may seem to be fairly straightforward for most people who interacted with the web based resources We simply know whether a particular link proposed by a search engine makes sense in relation to our intention With Magpie we suggest a strategy how such a sense making capability could be built into a web browser Before discussing the details of how this is accomplished let us mention what Magpie is and is not o Magpie is an enhancement to the standard browser It is not supposed to replace it o Magpie works with the explicit information that appears on the web pages and in various knowledge repositories It only associates explicit information at one end with a potentially relevant one at another place It does not read in between the lines or second guess the authors intentions o Magpie provides means for a structured and informed exploration of the web resources It helps with complex reasoning but is not omniscient Consider the following scenario as a typical example of how Magpie can be used to
15. make sense of a particular organization and its internal relationships Looking at a list URL http kmi open ac uk projects magpie Author Martin Dzbor KMi Page 2 of 14 of publications one member of the organization published on the web we can guess a few possible relationships hidden on such a page For example it shows the member s community of practice people s he interacts with topics and issues s he specializes in and similarly However what about information such as what specific projects 1s that member involved in What generic interests does that member have What about other collaborators and colleagues in addition to the paper co authors System requirement for Magpie Magpie is a plug in written for Microsoft Internet Explorer and this fact determines most of the requirements your system must satisfy to use Magpie and thus semantic browsing effectively e Operating system MS Windows Win32 platform this is due to the fact that Magpie is shipped as DLL which only runs on Win32 systems We have tested the plug in on the following operating systems Windows 2000 Pro Windows XP Home and Pro e Web browser MS Internet Explorer releases 5 5 or higher We strongly recommend using currently most up to date nternet Explorer 6 e Hard disk space At least 20MB free space on your C drive e Connection to Internet This is not needed 1f you wish to only highlight the ontological entities on off line web pages Ho
16. niversity 6 0 spi Oj x File Edit view Favorites Tools Help Back gt amp A A Reach Gyravorites meda lt 4 B Sp A S BH 4 Links si A Magpie Person E Project i Research Area E Organization x KNOWLEDGE MEDIA John Domingue INSTITUTE gt I A John s Home Page ABOUT US NEWS PEOPLE e Members 44 0 1908 655014 Contact John Domingue Keys KMi Deputy Director Semantic Web Semantic Web Services Organisational Learning Knowledae Modelling Software Visualisation Related Research Areas Affiliates PROJECTS PUBLICATIONS SOFTWARE 5 7 T Ba Relevant Publications I m KMi s Deputy Director and I m interested in how semantig web and can support the creation and sharing of knowledge In partigular at pr Related Stories in how the ontological enrichment of web pages can suppor interpreta Related Technologies interested in how semantic web services can be easily creafed configu Related People Related Projects Martin Dzbor Martin s Home Page 44 0 1908 858524 Contact Martin Dzbor Keys ClimatePrediction Net Jabber Design theory Knowledge modelling Reflection and Framing EE E Semantic Web Figure 6 Web page with entities of two categories annotated and semantic menu invoked over phrase Knowledge Management gt Note that the style may differ if you choose a different scheme or fonts in your Windows settings Author Martin
17. number of instances and of course computer power 7 Ifthe loading and lexicon initialization is successful you should see Magpie toolbar changing its appearance and show the top level classes categories from a particular ontology Each class is a button with a different colour label For details of the screen see Figure 2 Fie Edit view Favorites Tools Help 4 Bak gt amp at Gisearch Favorites iMedia C4 D 3 Aagi Links A Magpie _ Ferson x eroiect ki _ Fesearch Area Organization Change Button Text Change Highlight Colour A Figure 2 Magpie toolbar after loading a lexicon and showing the top level classes Having come so far means that the Magpie extension is properly set up and ready to be used for browsing Now you may start browsing the web by typing a URL web address in your web browser or by searching your favourite web engine We shall look at the actual behaviour during browsing in the next section However let us mention two simple configuration changes you can perform at this stage first First you can change a button label for each of the top level classes categories by using the first option Change Button Text in the pull down menu shown by pointer A in Figure 2 To access the pull down menu you click on the little arrow next to the button you want configure and then choose the option you want Similarly you may want to change the colours used by Magpie to highlight entit
18. oke a suite of semantic services contextually relevant to phrase John Domingue from category Person Right click on any highlighted occurrence of John Domingue and menu shown in Figure 5A should appear Choose menu option Works With Translated into human friendly language this semantic service requests a list of all people with whom John collaborates The answer is given in Figure 8A Finally let us choose option Shares Interests With for the same instance John Domingue Effectively we are asking who are the other people in KM1 with interests similar to those of John This is one of services requiring more complex reasoning Hence it may take a few seconds for the answer to show in the pop up window of your web browser Nonetheless the expected answer shall resemble that in Figure 8B As we can see from the two figures some of the responses of the ontological reasoning engine come back as a plain text see Figure 7 Other response especially those related to the people or other entities that have their own web sites come back as URLs You may therefore click on any of the people listed in Figure 8 Even better click and drag one of the URLs to your browser window This trick will leave Magpie toolbar loaded and available to annotate personal pages of KMi members Related and future functionalities In addition to the basic functionality of highlighting ontological entities and providing addition
19. ologies Knowledge Modelling ictonia Uren Semantic Web Area Knowledge Modellmg Murray Altheim Ontologies Knowledge Modellng http plainmoor open ac uk 3000 Fems m E4 Knowledg e Management E E SE EA gt a Matej Koss Knowledge Reuse Ontologies Knowledge Modelling Knowledge Management A Magpie Lukas Trejtnar Knowledge Publishing Knowledge Reuse Knowledge Maintenance Ontologies Knowledge Modelling Knowledge Management Organizational Learning John Domingue works with Bertrand Sereno Semantic Web Area Human Computer Interaction e Peter Scott e Victona Uren e Zdenek Zdrahal e Mare Eisenstadt e Paul Mulholland Yanna Vogiazou Human Computer Interaction Clara Mancini Human Computer Interaction Simon Buckingham Shum Knowledge Publishing Knowledge Reuse Semantic Web Area Ontologies Knowledge Modelling Incidental Ka Knowledge Management Organizational Learning Dynamic Linking Human Computer Interaction Paul Maceachern Artificial Intelligence Research Area Marut Hasan Knowledge Reuse Knowledge Management Artificial Intelligence Research Area A B e Bertrand Sereno e Marut Hasan e Simon Buckingham Shum e Martin Dzbor e Enrico Motta Figure 8 Two ontological responses related to the instance John Domingue A Who does John work with and B With whom does John share research interests Author Martin Dzbor KMi Page 10 of 14 For slightly more sophisticated reasoning we may now inv
20. or gs Martin s Home Page 44 0 1908 858524 Contact Martin Dzbor Keys ClimatePrediction Net Jabber Design theory Knowledge modelling Reflection and Framing Knowledge Management Semantic Web x Figure 3 Web page after clicking on button Person i e highlighting people in the page who are known in the loaded ontological lexicon Org Intranet Knowl Management onto Concept highlighting As we mentioned above we explain the highlighting functionality on KMi People s web page Assuming you already navigated to the appropriate URL and loaded Org Intranet Knowl Management onto lexicon start with clicking on button labelled Person on the Magpie toolbar see pointer A in Figure 3 This button instructs Available in several formats from http www aktors org publications ontology http kmi open ac uk people Author Martin Dzbor KMi Page 7 of 14 Magpie to find any phrase on the page that might be considered as a KMi member By default the button highlights concepts with light green colour see snapshot in Figure 3 As could be expected Magpie found its authors and highlighted them in green Before we start exploring the semantic menus and associated semantic services we can highlight other perhaps less obvious concepts and phrases on the web page For example to annotate known research areas click on the appropriate button In this case it would one with a sky blue icon and label Research Area
21. requested services we may easily find out more information about a person e g John Domingue However using trigger services these projects can be easily listed next to a browsing window see Author Martin Dzbor KMi Page 11 of 14 pointer A in Figure 9 Moreover each project pushed by a trigger services is also associated with a URL of the page which helped trigger a particular response This enables the user to make semantic bookmarks rather than syntactic In other words remembering page that is relevant to a particular concept phrase or topic rather than recording a URL without the context it was accessed From Figure 9 we can see that other projects that are related to John though not mentioned explicitly on the currently viewed web page include e Lyceum Meet O Matic and KMi Stadium these are Marc Eisenstadt s contributions to potentially relevant projects e BuddySpace climateprediction net and Magpie Martin Dzbor s contribution e Patman WebOnto PlanetOnto etc John Domingue s contribution Obviously there are some overlaps in this list for instance climateprediction net or Magpie involve all four individuals listed in box B thus each of them would contribute with these particular projects as potentially relevant ioi x eres ol x 4 5 48 5 8 4 E KNOWLEDGE ti a Dea REUSABLE COMPONENTS FOR C O x KHNOWLE DGE oon E o a n a c u k MODELLING r E tr
22. shed successfully a window with a respective system message should appear on your screen In case that Magpie registration with the system was not successful the most typical reasons are e You don t have sufficient privileges to install register a new application Resolution Please log in as Administrator or ask your system administrator for assistance e You have an Internet Explorer application open and system cannot modify its settings while running Resolution Close all windows of Internet Explorer and re run the above command Since Internet Explorer occasionally stays in the computer memory as a process with no visible windows showing it may be necessary to have a look at the running processes too e I seeno Internet Explorer windows on my desktop what about the processes e Resolution Right click on your system bar the one where Start button 1s and in the menu choose option Task Manager In the Task Manager window go to tab labelled Processes find one called IEXPLORE EXE and highlight it Next click on button End Process and confirm it These are the most common reasons for failing to operate Magpie extension properly If you encounter anything else please let us know To test whether Magpie really is in your web browser launch Internet Explorer Once it starts up you should see a little button usually towards the right hand end of the Internet Explorer s button bar
23. t al 2003 Semantic Layering with Magpie Handbook on Ontologies in Information Systems S Staab and R Studer Springer Verlag Dzbor M J Domingue et al 2003 Magpie Towards a Semantic Web Browser Proc of the 2nd Intl Semantic Web Conf Florida USA Gruber T R 1993 A Translation approach to Portable Ontology Specifications Knowledge Acquisition 5 2 199 221 Author Martin Dzbor KMi Page 14 of 14
24. t with Magpie and any web page enrichments facilitated by Magpie extension e First the user may highlight concepts of any particular class category shown in the Magpie toolbar that happen to appear on the web page e Second for each highlighted item or phrase the user may invoke a semantic or contextual menu for browsing the value added information about the instance 3 Members Microsoft Internet Explorer provided by The Open University 6 0 spi a ioj x File Edit View Favorites Tools Help d Back gt amp A A Asearch GgFavorites Meda lt 4 D S Hi S a Links z A Magpie Person B Pl Project 7 Research Area Gi __ organization z KNOWLEDGE MEDIA John Domingue INSTITUTE A John s Home Page ABOUT US NEWS 44 0 1908 655014 2 PEOPLE Contact John Domingue e Members Bl Affiliates Keys KMi Deputy Director Semantic Web Ontologies Semantic Web Services Organisational Learning Knowledge Management Knowledge PROJECTS Modelling Software Visualisation PUBLICATIONS I m KMi s Deputy Director and I m interested in how semantic web and knowledge technologies SOFTWARE can support the creation and sharing of knowledge In particular at present I am interested in how the ontological enrichment of web pages can support interpretation I am also interested in how semantic web services can be easily created configured and deployed e amp a ss a Martin Dzb
25. uch as RDF for lexicon definition Similarly we will try to offer a web service creating a Magpie compatible lexicon from any arbitrary ontology submitted as an input Nonetheless these are future extensions In order to test your Magpie you need to load the lexicon first Assuming you have downloaded installed the default Magpie application follow these steps Author Martin Dzbor KMi Page 5 of 14 1 Click on Magpie button to show the Magpie toolbar in your web browser You should see a situation similar to that in Figure 1 2 Next click on the little arrow next to the label Magpie on the Magpie toolbar see pointer B in Figure 1 A menu similar to that depicted in the figure should appear 3 Choose the first option labelled Load Ontology and then From local file A file selection dialog window appears letting you choose the lexicon with onto extension 4 Navigate to the location folder where you stored the lexicon usually it would be found in C Magpie or Ontology Store sub folder of the main installation directory mostly C Magpie 5 Choose Org Intranet Knowl Management onto for the purposes of interacting with a knowledge management application of Magpie 6 Clicking on Open button Magpie starts loading lexicon Note that this may take a while around 2 5 seconds for Org Intranet Knowl Management onto lexicon depending on the size of the lexicon
26. wever Internet link is essential if you want to take advantage of ontology supported semantic browsing and semantic services Magpie has been tested on all systems and applications mentioned above Please inform us if you find additional operating systems or IE versions where it works Magpie as a Semantic Web browser The background information mentioned in the previous section is essential for semantic browsing In case of Magpie this information is stored by means of ontologies Gruber 1993 which are formalized explicit representations of domains and or problem matters In order to provide the full functionality the user has to have access to one or more ontologies s he intends to use to make sense of browsed web pages To achieve as high flexibility as possible Magpie has been designed modularly This means that only a relatively small part resides locally in the user s web browser Majority of the functionality is provided remotely by accessing various semantic services knowledge based reasoning tools and data over network Two standard parts of Magpie needed for a basic operation include o Magpie toolbar loadable into a web browser this is delivered as a dynamically loaded library DLL o Ontology based lexicon that can reside on the local computer e g user s private representation of a problem domain We are currently developing functionality that would enable users to subscribe to shared public ontologic
Download Pdf Manuals
Related Search