here - Newcastle University
Contents
1. Free text input question similar as above except that the entered answer can be any text e g Enter the name of the largest ocean on earth 4 Newcastle University E Voting Client Please enter the Session ID and Passcode of the election you wish to vote in Election Session ID Passcode Submit View Receipt History View Results Fig 2 Login screen for the voting client The system is only verifiable when the voting questions are of the first two types our verifiable voting protocol requires knowing the names of the candidates before the election However we still support the latter two types of questions as we consider them useful features even though the voting results cannot be verifiable in the cryptographic sense When a coordinator creates a voting session there are a few options he needs to configure Whether using a group passcode or individual passcodes The maximum number of students in the class denoted as NV The auditing factor F which allows each student to audit a vote up to F times by default F 5 The security level L bits by default L 128 The length of the receipt R characters by default R 5 A group passcode is a single passcode available to all students in the class In this setting the teacher informs students of a session ID and the group passcode which are needed to log into the particular voting session see Figure 2 However one drawback wit2. ac uk At the time of the trial the iPhone app was still under development so iPhone Stage 2 receipt for cancelled Yes Cancellation Check receipt at Your vote has been cancelled You can either skip or retry this evoting ncl ac uk receipt question Session ID 96 Question ID 1 Ballot ID 0 Er Receipt UKRGO AXFEJ Zz Attempts remaining 4 a Case A user chose cancel b Case A receipt for cancel lation lation Stage 2 Receipt i Check receipt at Confirmation evoting ncl ac uk receipt Your vote has been successfully registered Session ID 96 Question ID 1 i l Ballot ID 1 ose Receipt Receipt QJND6 3VSRP c Case B user chose confir d Case B receipt for confir mation mation Fig 4 Stage 2 voting interface and receipt lt Newcastle University Results 50 25 25 Refresh Fig 5 Display of tallying results users had to use a web interface to vote At the end of the workshop we received 26 completed questionnaires among which 17 participants voted through the Android app and 9 through the web interface The feedback questionnaire consisted of 10 statements and respondents were asked to indicate their agreement or disagreement on a Likert scale from 1 to 5 i e strongly agree agree neutral disagree and strongly disagree The statements were as follows 1 Joining a new session was easy 2 I unders
3. the stage 2 receipt for the cancellation case Figure 4b A stu dent can repeat the same cancellation operation up to F times recall that the value F is configurable On the other hand if the student chooses to confirm the interface will show the vote has been casted Figure 4c together with a stage 2 receipt for the confirmation case Figure 4d Same as before to verify the stage 2 receipt the student simply needs to check that the content on the receipt matches that published on the public bulletin board This requires no knowledge of cryptography As long as all receipts are available on the public bulletin board anyone with cryptographic knowledge and computing skills will be able to verify all receipts in a batch The third and last phase is the tallying process When all students have casted their votes the coordinator would end that particular voting session through a web interface The voting results are immediately available Figure i Newcastle University 1 Does E T exist Select one of the following options Stage 1 Receipt Confirmation Check receipt at You have entered Yes evoting ncl ac uk receipt To register your vote select Confirm If you wish change your vote select Cancel a Initial voting interface b After choosing Yes c Stage 1 receipt Fig 3 Stage 1 voting interface and receipt 5 shows an example of the tallying results in a bar chart The same results
4. Verifiable Classroom Voting Where Cryptography Meets Pedagogy Feng Hao Dylan Clarke Carlton Shepherd School of Computing Science Newcastle University feng hao dylan clarke c g shepherd ncl ac uk Abstract In this paper we propose and have implemented the first verifiable classroom voting system The subject of secure classroom voting has so far received almost no attention from the security com munity Though several commercial classroom voting systems have been available none of them is verifiable State of the art verifiable voting pro tocols all rely on finding a set of trustworthy tallying authorities who are essentially cryptographers and computer experts in the first place and hence are completely unsuitable for classroom voting Our system design is based on self enforcing e voting a new paradigm that was first presented at SPW 12 Hao Randell and Clarke A self enforcing e voting scheme provides the same End to End E2E verifiability as other e voting schemes but without involving any tallying authorities The removal of tallying authorities brings several compelling advantages in real world voting scenarios here classroom voting is just one exam ple We have piloted the use of the developed verifiable classroom voting system in real classroom teaching Based on our preliminary trial expe rience we believe the system is not only scientifically valuable but also pedagogically useful 1 Int
5. are also available on the voting website together with all receipts i e audit data We provide an open source Java program on the voting website to facilitate any interested party to cryptographically verify the integrity of the results based on the audit data 3 Trials 3 1 Usability trial We conducted a voting trial workshop at the School of Computing Science Newcastle University on 3 September 2012 The participants were mainly MSc students who had just submitted their dissertations We provided pizza catering for all participants With this workshop we aimed at three birds with one stone to trial our newly developed verifiable classroom voting system to serve as a farewell party for MSc students as many of them would leave shortly and finally to give some MSc students a chance to present interesting results in their dissertation projects and let all participants vote for their favorite presentation using the classroom voting system There were in total around 40 participants who were mainly MSc students Five students presented their dissertation projects and afterwards we asked all participants to vote for the most entertaining presentation In this case the integrity of the voting result must be ensured so we used the individual passcodes scheme as described in Section 2 During the trial participants had two ways to vote 1 using an Android app 5 version 1 0 0 2 using a web interface at http evoting ncl
6. ave developed two voting clients an Android app 5 and an iOS app 4 to support voting from iPhone iPad Android phone and Android tablet In addition we provide a generic web voting interface so people with any other types of smart phones e g windows 8 blackberry etc or a laptop can still vote as long as the device has a web browser and is connected to the Internet 2 2 System configuration There are three roles involved in the use of the system administrator coordinator and voter The administrator is responsible to maintain the availability of the web server A coordinator usually a teacher is someone who coordinates voting in a classroom The system can accommodate many coordinators at the same time Finally voters are usually students in a class Classroom voting is arranged according to voting sessions A voting session consists of a list of voting questions We support four types of questions in the system 1 Single answer question students can only choose one answer e g which is the largest country in the world A Russia B China C America D India 2 Multiple answer question students can choose multiple answers e g which of the following countries are members of Commonwealth A Singapore B India C Austria D Canada 3 Free numeric input question there are no given answers and students are free to enter any numeric value e g Enter the value of 7 to the two decimals 4
7. h all important security elements accounted at the outset of the design the public confidence in the new system and their acceptance will gradually grow Question ames Agree oe a eras ac eda Average score sree option DO Co a po Stromigly agre sz is s i 0 0 LA Strongly agree 7 an a o 0 23 Strongly agree sas 3 0 ObUr a e 4 uy s 3 0 227 Agree 7 a ao fs gt 275 Neutral sf 0s 0 sUr 9 n ols i 1 196 Agree o s dP fof 3 o Neutral Table 1 Summary of the 26 received questionnaire answers 3 2 Pedagogical trial Following the success of the usability trial in September 2012 we made several improvements to the Android app to make it more user friendly Also we pro vided an iOS app 4 for those using iPhones and iPads to vote In October and November 2012 we first trialed the system in real classroom teaching on the Cryptography BSc final year and System Security MSc first year mod ules in which the first author is the module leader On 10 January 2013 at the last revision lecture of Cryptography the first author prepared ten revision questions for the class gave students 15 minutes to discuss among themselves and asked them to vote for the best answers At the end of the lecture a stu dent survey was conducted using the same voting system to collect the feedback The survey questions and the tallied answers within brac
8. h this authentication mechanism is that one student can vote multiple times by re using the same passcode In many circumstances this is not an issue as there is no incentive for students to double vote However in some cases when voting involves sensitive questions such as rating a lecturer s performance a group passcode would be inadequate Individual passcodes should be used instead In the individual passcodes scheme each student is assigned a unique pass code The web server first generates N random passcodes recall that N is the maximum number of students in the class The coordinator then prints out all N passcodes each on a paper slip The paper slips are physically mixed up in front of the students before being distributed to the class One student can only take one passcode After voting is finished the public bulletin board will show how many passcodes have been used This number should be matched to the actual number of students in the class e g based on a signed class attendance sheet Any significant discrepancy would suggest something wrong e g ballot stuffing which demands further investigation 2 3 Voting protocol To implement the system we adopt the Direct Recording Electronic with In tegrity DRE i protocol 7 which is under the category of self enforcing e voting protocols 6 The DRE i protocol provides the same end to end veri fiability as other verifiable voting protocols but without requir
9. ing any tallying authorities The protocol has three phases setup voting and tallying The setup phase involves pre computing cryptograms for all electronic ballots as specified in 7 Depending on the size of the class and the number of questions this phase usually takes several minutes to complete The second phase is voting Figure 3a shows the initial voting interface of the Android app for a single answer question To cast a vote a student follows two stages 1 selecting an answer 2 confirming or canceling the previous selection In the first stage the student makes a selection let us assume he selects Yes In the next interface the app shows that Yes had been selected and asks the student to Confirm or Cancel Figure 3b There is also a third button Receipt which leads to the display of a stage 1 receipt Figure 3c The student can verify the receipt by checking that the same content on the receipt has been published on the public bulletin board a publicly accessible website The second stage handles the student s choice of confirm or cancel Sup pose the student chooses to cancel essentially this is to perform voter initiated auditing 7 The voting interface will show that the previous selection has been canceled Figure 4a the student can proceed to the next question if any or re try the same question There is also a button Receipt which leads to the display of
10. ity to par ticipate in a demo of the TurningPoint system at Newcastle University This research work was motivated by that experience However a notable limitation with TurningPoint and in fact all existing classroom voting systems is that the voting results are not verifiable There is no means for students to check if their votes have been recorded and tallied correctly The integrity of the results may be affected by many factors e g hardware malfunction of the voting device lost signal in the radio frequency transmission software bugs malicious attacks where an adversary tampers with the back end software to arbitrarily modify the results One might question why we should care about the verifiability at all if the tallying results turn out to be wrong it is probably not too big a deal After all the classroom voting result is not as sensitive as that in political elections How ever we believe verifiability is still important First of all it provides confidence on the accuracy of the tallying results If any hardware failure or a software bug causes the tallying procedure to go astray the error in the result will be caught publicly if the system is verifiable Second though classroom voting questions are usually not sensitive there are exceptions for example when the system is used as a module assessment tool to rate a lecturer s teaching performance By taking security into consideration in all conditions we can
11. kets are summarized below Question 1 Does the voting make the lecture more fun Answers Yes 26 No 2 Question 2 Does the voting help you learn Answers Yes 26 No 2 Question 3 Do you find it useful to have a small group discussion before voting Answers Yes 25 No 3 Question 4 How do you think that the amount of voting used in this lecture should change Answers More 10 Less 1 Remain the same 16 Question 5 Do you recommend classroom voting for teaching the same module next year Answers Yes 26 No 1 The survey results clearly indicate the pedagogical value of the developed class room voting system The vast majority of the students in the class found the classroom voting system quite fun to use We believe fun is a critical factor in learning by making learning a fun process we are able to better retain the students attention in the class and improve their learning outcome It is also worth noting that we used about a quarter of the time in a lecture 1 hour for classroom voting We were initially concerned if that was too much But based on the feedback 16 out of 27 expressed that was an adequae percentage another 10 nearly one third students actually wanted more voting in the class This is further evidence to show that students generally liked the system 4 Conclusion In the paper we have presented a pioneering classroom voting system that is verifiable This sys
12. make classroom vot ing more widely useful Finally by making the system verifiable students will have an opportunity to learn and practise the fundamental trust but verify principle in routine classroom voting This can prove relevant when they later participate in more serious national elections Besides a lack of verifiability there are other limitations with the exist ing classroom voting systems They generally use proprietary devices as voting clients This however not only imposes vendor lock in but also creates serious logistical issues simply transporting the physical devices into and out of the classroom can be a laborious task In addition they require installing a pro prietary receiver in the classroom This seriously limits the portability of the system as voting is confined to only designated classrooms 2 System design In this section we will propose a verifiable classroom voting system and show a concrete implementation Our system addresses all the problems we explained above http www iclicker com http www turningtechnologies co uk 3 http www einstruction com i Web server i A ieee Android om 4 tablet iPad Android phone Fig 1 Verifiable classroom voting system using mobile devices as voting clients 2 1 Overall architecture Figure 1 shows the overall architecture of our system At the client side students use their own computing devices to vote We h
13. roduction Classroom voting is a powerful new pedagogy which was first developed for the physics classroom by Harvard University s Eric Mazur in his influential book Peer Instruction A User s Manual 1 and subsequently extended by other academics to teaching mathematics and other subjects 2 In this teaching technique the teacher first poses a set of multiple choice questions to a class of students gives them a few minutes to discuss in small groups and asks them to vote for the best answers Typically a student submits the vote using a special hand held device known as the clicker 2 3 that sends radio frequency signals to a special receiver installed in the classroom The receiver tallies votes in real time and displays the results over a projector providing instant feedback to the students and the teacher alike Several studies have reported success of using this technique to retain the students attention to The work was supported by Newcastle University Innovation Funds and partly by the ERC Starting Grant SEEV No 306994 increase the classroom interactions and to improve the student learning outcome 2 3 There have been several commercial classroom voting systems available e g iclicker TurningPoint and eInstruction In particular the TurningPoint vot ing system has been adopted and trialed by a number of universities in the UK including Newcastle University The first author had an opportun
14. tem serves as a good example to demonstrate the power of the underlying self enforcing e voting paradigm Through putting the system into the real classroom teaching and collecting the student feedback we show that the system has also demonstrated great pedagogical potential to enhance the students learning experience in a traditonal classroom environment References 1 E Mazur Peer Instruction A User s Manual Prentice Hall Series in Educational Innovation NJ 1997 2 K Cline and H Zullo Teaching Mathematics with Classroom Voting With and Without Clickers Mathematical Association of America 2011 3 D Bruff Teaching with Classroom Response Systems Creating Active Learning Environments Jossey Bass 2009 4 Link to the iOS app for the verifiable classroom voting application https itunes apple com us app id565080670 5 Link to the Android app for the verifiable classroom voting application https play google com store apps details id uk ac ncl evoting 6 F Hao B Randell D Clarke Self Enforcing Electronic Voting Proceedings of the 20th Security Protocols Workshop SPW 12 Cambridge UK 2012 7 F Hao M N Kreeger Every Vote Counts Ensuring Integrity in DRE based Voting System School of Computing Science Newcastle University Technical report No 1268 2012
15. tood how to join a new session 3 I understood how to answer questions 4 Answering questions was easy 5 I understood how to check the receipt 6 I understood why I might want to check the receipt 7 I felt confident that my answers had been recorded correctly 8 I understood how to view the results 9 Viewing the results was easy 0 10 I felt my answer was sent anonymously We summarize the received 26 questionnaire answers in Table 1 There was no obvious difference in the answers between those voting through the Android app and those through the web interface so we combine all answers in one table In general the feedback was encouragingly positive Participants generally found our verifiable voting system easy to use see Table 1 However some people expressed neutral opinions about the security of the system Despite that we designed the system to be verifiable and we physically shuffled the passcodes to ensure anonymity roughly half of the participants indicated they were not sure whether the vote had indeed been correctly recorded and whether the voting was anonymous These are useful lessons which teach us that e voting is not only a security problem but also a subject of of psychology and voters perception of security We do not believe anyone should immediately accept a new voting system just because it is verifiable or has security proofs But we do believe that given a verifiable voting system wit
Download Pdf Manuals
Related Search