ONLINE EXAMINATION SYSTEM USING ARTIFICIAL INTELLIGENCE

International Research Journal of Engineering and Technology (IRJET) e ISSN: 2395 0056

Volume: 09 Issue: 06 | June 2022 www.irjet.net p ISSN: 2395 0072

ONLINE EXAMINATION SYSTEM USING ARTIFICIAL INTELLIGENCE

1Asst. Professor, Dept.of Information Science and Engineering, Canara Engineering College, Karnataka, India 2 3 4 5Student,Dept.of Information Science and Engineering,Canara Engineering College,Karnataka,India ***

Abstract Online Examination System is an application builttomodernizetheexaminationprocessthathasbeena major challenge, especially during the covid pandemic. Remote proctoring is one technology which is trending in the era of social distancing. For a couple of years, e learninghasbecomefamousasa resultofitssimpleclient intelligence. All things considered; the significant danger looked at by the examination local area is the procedures that are being utilized. We have utilized equipment, for example, a web camera to catch sound and video along with dynamic window catch. This blend shapes the contributionstoa decide basedsurmisingframeworkthat can conclude whether any misbehaviors have occurred or not The software will be able to successfully conduct the exam virtually without the physical presence of an instructor.Theexamsaretime basedandsecuresothatall thecandidatesgetafairchancetogivetheexamination.In this application, we present a strategy to stay away from the actual presence of a delegate all through the test by making an exhaustive performance of multiple tasks frameworks.

1. INTRODUCTION

ArtificialIntelligencealsoknownasmachineintelligence,is abranchofcomputersciencethatfocusesonmanagingand building technology that can learn to instantly make decisions and carry out actions instead of a human being. AI based Online Examination Proctoring uses AI driven algorithms for user identification and cheating identification.Itwillmakeanextensivereporttoassistyou withaffirmingontheoffchancethataninternetbasedtest has been finished with machine driven invigilation delegating. Simulated intelligence fueled remote administeringcanguaranteethestarttofinishsecurityfor Online Exams and keeps the interaction from acts of neglect and cheating. Because of the COVID 19 pandemic, the whole world is following a major downfall in terms of economy, day to day activities are being restricted due to lockdown!Remoteproctoringisthetechnologythatallows ustomonitoronlineexamsforcandidatesgivingtestfrom different locations,anywhereonthebackoftheearth.The ubiquity of online and electronic programming has filled

quicklyasoflate.Themannerinwhichwedooureveryday exerciseshaschanged.Throughoutrecentyears,theuseof Internet andcomputers as educational apps hasincreased quiet quickly There are different factors of remote proctoring such as video proctoring, image proctoring, audio proctoring, monitoring the screen of the user taking the online exam. With the help of all these functionalities, we can securely conduct Online Exam, that too under a secure, monitored, and cheat free environment. Face recognition is the key component of the verification process for remote users during the online exam process. Thesystemtakespicturesoftheusertakingtheexamand it is compared with a photo stored in the system at the startoftheonlineexam.Anytimeoftimeduringthetest,on theoffchancethatthereisabungleordeviationintheface or more than one face recognized before the presentation ornooneelse'scountenancesidentifiedbeforethecamera duringtheweb basedtestthentheframeworkrecordsasa caution Artificial intelligence based remote administering likewise denies the client from exchanging the windows while showing up for the booked Online Examination SystemUsingAItakescareofwithitsextralightdelegating componentforthesakeofImage basedadministering.The mechanismimposedbytheproctoredexaminationprocess enforces this restriction to keep the Online Exam environment under control. Image Proctoring provides secure environment for online exam prominently and the demand in the technology is rapidly increasing due to severalfactors.

2. RELATED WORK

Theproposedsystemaimstoestablisha structurethat meets the requirements of the examination process in universitiesandhighereducationinstitutions. Recordsare a test, and it requires greater investment and work to accomplisharrangingofrecords.Toperformonlinechecks, makinga differentapplicationisn'trequired.Theessential inspiration is to settle the troubles that might exist in manualtestingframeworkslikethebaninthereadinessof tests Remoteexaminationandproctoringaresignificantly gaining importance in the wake of the accommodation of comfort, security, and accessibility. This will not just increase the importance of course or remote based examinationsbutitalsohelpsinMOOCs(massopenonline courses) and other credit based certifications for the

International Research Journal of Engineering and Technology (IRJET) e ISSN: 2395 0056

Volume: 09 Issue: 06 | June 2022 www.irjet.net p ISSN: 2395 0072

concern of establishing credibility. The proposed online examination software the monitoring system uses advanced, secure, and reliable Artificial Intelligence to monitor the students and review the examination. Web based automated examination system which detects all unusualactivitiesandflagsthem,toensurefairproposition ofexams.Invigilationoftheexamisnotrestrictedtoafixed time and actual test grounds anylonger. This has revoked aninterestintheassortmentofonlineremoteexamination arrangementsandsupportingtechnologicaladvancements.

In this application, we present a strategy to stay away from the actual presence ofa delegate all through the test by making an exhaustive performanceof multiple tasks frameworks. This assessment framework utilizes fewer assetsandlessenstherequirementforQ&Acontents,test room booking, organizing invigilators to test corridor, planningwithinspectors,andthat'sjustthebeginning.The online examination system uses client/server architecture methodology. The proposed framework is simple and adaptable on account of future upkeep and improvement because every subsystem can be taken care of independently without effect on another framework. The customer by utilizing an internet browser can associate throughweborneighborhoodhavewiththeserverwhere PHPandMySQLintheserver sideareliableforthegetting ready tests cycles and save and return information from dataset.Itisessentialtomakesurethatthedatasetisvery much planned. The documents names picked to mark the tables made inside the information base endeavor to mirrorthetable'smotivationand,thusly,addtoawell plan framework. Utilizing an open source language gives us greater adaptability, and yet it required more opportunity tobemodified.

Thesystemforcedbythedelegatedassessmentprocess upholdsthislimitationtomonitortheOnlineExamclimate. Picture Proctoring has made it conceivable to conspicuously get the web based test conditions and the demand in the technology is rapidly increasing due to severalfactors.

Instructing and mentoring, exhorting and advising and data and organization are the institutional frameworks associated with student support. Student support is characterized as "every one of the exercises and componentsinschoolingthatreactstoaknownstudentor a gathering of students, which are intended to aid emotional and precise domains of the learning process. Students get themselves registered for several online programs and it's increasing exponentially. But students droppingoutofthesecoursesisalsomoreduetoimproper support services. As ODeL establishments grow their scholarly projects, their understudy populace becomes more assorted. The expanding variety of understudies featurestheneedtocreateandcarryouthelpbenefitsthat

ensure the understudies' achievement. OSP is an online application that offers a combination of organizations. It gives students reliable permission to information and organizations

In Оnline Exаminаtiоn System, reseаrсhers аre attempting tо change thetrаditiоnаlmаnuа exаminаtiоn in the eduсаtiоnаl institutiоn. The оnlineexаminаtiоn system is effeсtive in time mаnаgement.In this рарer, we рrороsed а system with аn аutоmаtiс аssessment teсhnique. This аlgоrithms is tо determine the frequenсy оf wоrds, Keywоrdmаtсh ,linguistiс аnаlysis, аnd grаde generаtiоn. The system is imрlemented with the helроf MySQL аndРhрStrоm. Оurаdvаnсed system is аbletоevаluаte the resроnse оf desсriрtive questiоn. Mаrk аllосаtiоn deрends оn this рerсentаge ассurасy оf the аnswer. The whоle рrосess соnsists оffоursteрs.Itwillmakea farreachingreportto assistyouwithaffirmingintheeventthataweb basedtest has been finished with machine driven invigilation delegating. AI powered remote proctoring can ensure the end to end security for Online Exams and prevents the process from malpractices and cheating They аre Extrасtiоn оf keywоrds, weightаge оf keywоrds, Feаture mаtсhing,Sсоregenerаtiоn.Mоstоftheсоmрetitiveexаms аreсоnduсtedоnline.Beсаusetheyаretime effiсient, соst effiсient, аnd seсure. In the rural аreа, students have to travelfromfaralongtogiveexaminationsintheirallocated center.Toconclude,learnersareanimportantaspectofthe ODeL environment. Hence their welfare needs to be the maingoalofthesystem.

3. SYSTEM ACTIVITIES REQUIREMENTS

Itispossibletoimagineandcombinethepartsoftheentire system in many different way. Ourproposed system, consistsofthe Administrator Module and the Student Module. The Admin Module consists of further two sub modulesAdmin and Instructor and the student module consistsofStudentandOnlineProctoring

Figure 1: Block diagram of online examination system using AI

International Research Journal of Engineering and Technology (IRJET) e ISSN: 2395 0056

Volume: 09 Issue: 06 | June 2022 www.irjet.net p ISSN: 2395 0072

4. SYSTEM DESIGN AND IMPLEMENTATION

Subsequent to laying out an association through the Internet or an intranet, the proposed web based check framework can work with the client/server design as displayed in beneathfigure

Figure 2: Dataflow diagram

Functional Requirement:  Studentscantaketheexamremotelywithout beingpresentintheexamhall.  The instructor will be able to monitor without havingaphysicalpresence.  Anymalpracticedonebyacandidatecanbeeasily identified.  Questions to the student can be provided randomlytoavoidcheating.  Real time Proctoring can be achieved to have a fairexamination.  Results and notifications to students can be providedinlesstime  Remote Proctoring empowers possibility to show up for an evaluation from a distant area while guaranteeingthehonestyofthetest.

Figure 3 Figure 4 Figure 5 Figure 6

5 EVALUATION RESEARCH METH0DOLOGY

The assessment research objective is to feature our assessment framework as a product for taking tests and dissecting the outcomes that we got from understudies. This product has the determination test strategy which is different decision test. In the wake of taking their tests utilizing our assessment framework, the understudies have been gotten some information about their analysis with this assessment framework. The span that this exploration has taken is 7 months, from October 2021 to June2022.

International Research Journal of Engineering and Technology (IRJET) e ISSN: 2395 0056

Volume: 09 Issue: 06 | June 2022 www.irjet.net p ISSN: 2395 0072

Our developed system is able to assess the answer of descriptive question. The whole process will be done via Artificial Intelligence. Allocation of marks depending on thepercentage oftheaccuracyexistintheanswer In this process, examiner will set a question also upload a model/standard answer of that question. After login an examinee will answer that question. Then the system will assesstheanswerofexamineebymatchingthekeywords. It will also check the linguistic analysis. After the evaluation it will calculate the score according to the correctnessoftheanswer.Thewholeprocessconsistsof4 steps.Theseare:

value in the assessment process. If it doesn’t happen then itwill allocatehighervalue fortheassessmentprocess.In thismodulekeywordextractionisdonebytwosteps:

1) Preprocessing: Reply from client might contain numerous superfluous words and images, for example, accentuation images, stop images, feelings images and so on. These are considered as commotion for the informationwhichwillbehandled

2) Keyword frequency generation: The system will do thekeywordfrequencygenerationafterthepreprocessing phase. It will classify all the keywords in this method and find the frequency ratio. He used the algorithm 1 for this operation. In this method, the array frequency set is initially null. Then it goes through the script of the entire answer and compares each word with the list. If it is not on the list, the word in the list is added as an index. The frequencyofthatwordwillincreaseifidentified.

Algorithm 1: AlgorithmforCalculatingWordFrequency

1Answer=Totalanswerafterpreprocessing;

2Arrayfrequency=null;

3 for Everywordsinanswer do

4 if WordisnotintheArray then

5 AddthewordasindexintheArray;

6 frequency[word]=1;

7 else

8 frequency[word]=frequency[word]+1; 9 end 10 end

3) Keyword: We used the algorithm 2 for the keyword comparison to test the answer script after calculating the keyword frequencies from both the model response and the user’s answer. The following algorithm primarily followsthreecriteriainthisprocess.Ifmorekeywordsare detectedthan model answer,then the marksare reduced. It also does the same thing in the answer script if it finds verylesskeywordsthanmodel answer. Withanadequate numberofkeywordsthataresimilartothemodelanswer, studentswillgetagoodscore.

Algorithm 2: AlgorithmforPatternMatchingUsingNaïve Algorithm

Evaluation Automation Module

By this module system will extract the keywords from boththemodelanswerandtheanswersubmittedbyuser. It is an important process for the whole system. If the repetition of keywords is very often it will be given less

1.SETK=1andMAX=S R+1.

2.RepeatStep3to5whileK<=MAX:

3.RepeatforL=1toR:IfTEXT[K+L 1]≠PAT[L],then: Go toStep5.

International Research Journal of Engineering and Technology (IRJET) e ISSN: 2395 0056

Volume: 09 Issue: 06 | June 2022 www.irjet.net p ISSN: 2395 0072

REFERENCES

[1] NeeleshChandraM[1],PiyushSharma[2],Utkarsh Tripathi[3], Ujwal Kumar[4], Dr. G.C. Bhanu Prakash[5]. AUTOMATING ONLINE PROCTORING THROUGH ARTIFICIAL INTELLIGENCE. Date of Issue:01|Jan2021

4) Grammar and Spelling: Finding the correctness of grammar and spelling is one of the main issues for assessingtheanswerscript.Wehaveusedthealgorithm3 for the method of linguistic research. It initially puts the linguistic analysis score at zero in this process. Then, in the answer document, it measures the number of spelling errors,grammaticalerrors,thetotalnumberofwords,and sentences. The number of spelling errors is then divided by the total number of terms and multiplied by 100 Similarly it divides the number of grammatical mistakes bythetotalnumberofsentencesandmultiplyby100and addboththevalues

Algorithm 3: LinguisticAnalysisAlgorithm

1Stringanswer=Totalanswerofthestudent;

2InitialLAScore=0;

3InitialSMistake=Numberofspellingmistakes;

4 Initial GMistake = Number of grammatical mistakes; 5 InitialTWord=NumberofWordsinanswer;

6InitialTSentence=NumberofSentenceinanswer;

7 LAScore = (SMistake) /(TW ord)∗ 100 + (GMistake/(TSentence)∗100

6. CONCLUSIONS

Usinganopen sourcelanguagegivesus morenoteworthy versatility, but then it anticipated that extra open door should be changed. The proposed Online Examination System can be actually embraced by schools and foundations to make the test more secure and more versatile. The system is divided into two essential subsystems that are expected to give the structure most outrageous benefit by means of circumspectly showing each subsystem administration. This structure centers around the popular evaluation system research at this point, arranging a lot of ordinary evaluations for the schoolstageandgivinganincredibleshapetofiguringout many tests, and have an unprecedented reference a motivationforvariouscollegesanduniversities.

[2] Muna R. Hameed [1], Firas. A. Abdullatif [2]. Online Examination System. Date of Issue: 3, March2017

[3] Percia V. SECRETO[1], Rhodora L. PAMULAKLAKIN. LEARNERS’ SATISFACTION LEVEL WITH ONLINE STUDENT PORTAL AS A SUPPORT SYSTEM IN AN OPEN AND DISTANCE ELEARNING ENVIRONMENT (ODeL). Date of Issue:July2015

[4] SumaiyaKabir[1],Md.ParvezHossain[2],Kaushik Mallik[3], Mansura Rahman[4], Md. Jahidul Islam[5], Ayesha Khatun[6] . An Extensive Online Examination System With Automatic Assessment Technique.DateofIssue:01,DECEMBER2019

[5] Aditya Nigam[1] · Rhitvik Pasricha[2] · Tarishi Singh[3] · PrathameshChuri[4]. A Systematic Review on AI based Proctoring Systems: Past, Present,andFuture.DateofIssue:24March2021

[6] SakirTasdemir[1],MehmetBalcı[2],AyseCabi[3] ,MustafaAltin[4],OsmanCabi[5].TheDesignand Application of Online Exam System Supported by Database.DateofIssue:15August2014

[7] Anjali Choubeya[1], AvinashKumarb[2], Ayush Ranjan Behra c[3], Anil Raj Kisku d[4], Asha Rabidase[5]andBeasBhadraf[6]AStudyOnWeb Based Online Examination System DOI: 01 April 2020

[8] Feras Al Hawari1[1], Mai Alshawabkeh[2], Haytham Althawbih[3] and Omar Abu Nawas[4], Integrated and Secure Web Based Examination ManagementSystem,DOI:10.1002

[9] Jegatha Deborah L [1], Karthika R[2] , Vijayakumar P [3] and Bharat S. Rawal [4], Yong Wang[5], Secure Online Examination System for e learning,DOI:02October2019

[10] Richa Bhardwaj [1], Online Exam Management SystemUsingDeepLearning,Issue4April2020