An Intelligent Question Paper Generator using Randomized Algorithm

 

An Intelligent Question Paper Generator using Randomized Algorithm

Parth Shah*, Uzair Faquih, Rutuja Devkar1, Yogesh Shahare1

1MGM College of Engineering & Technology

Abstract:- This is a problematic era because of the boom in the subject of advanced technology and the demand we are dealing with nowadays. Therefore, examinations ly ruil rle in heking ut n individual’s performance, nd rertin f exm questin ers hs nsistently been mtter f interest. And this is why it’s far critical to have an intelligent system for the growth of students in addition to checking their learning skills, thereby retaining a check on student performance. Faculties generate various question papers in keeping with the university’s assessment requirements. It’s very challenging for the teachers to make question papers with multiple questions that meet the course’s learning targets. We have rsed n utmted ress f question er generation tht isfast, streamlined, randomized, and secure. Every action executed task by the system is automated, so that storage space, bias, and security are not a challenge anymore. It automatically creates a variety of sets of questions now and then without worrying about replication and duplication from the initial exam at the same time as the question bank keeps growing.

Keywords: Paperless, randomization, automation, assessments, question taxonomy.

1. INTRODUCTIONIn today’s modern, ambitious world, an exam plays a vital role in checking the educational improvement of students, and the technology of the information era is now substituted through the productive application of the technology. Any product which can correctly reduce time and power consumption is accepted and preferred. So producing software from knowledge is a crucial task to do. In all the academic courses that reject a variety of tests, the instructor intends to create important documents in accordance with the guidelines and assessments of the autonomic university. It is much more challenging to deal with all the course features for teachers and avoid duplicating questions from subsequent estimates. There is no systematic process, and for this reason, this problem’s paper quality is predicted for individual teachers and qualifications.The truth that there is a shortage of experienced teachers makes the situation even worse for specifying courses, semesters, syllabus, and patterns. At times, all these factors might also deteriorate the quality of the question paper. The researcher says a good questionnaire is the right combination of subjects (questions) guided by various parameters: cognitive level, difficulty and distribution of scores on the questionnaire. Creating a good questionnaire that contains many questions related to gaining knowledge about the purpose of a course in terms of content and cognitive level is a difficult task. So, we are presenting an Automatic Question Paper Generator System which could reduce time intake by replacing the traditional

   approach of question paper generation. There re rvisins t enter nd edit dt suitble t ny edutinl rgniztin with mlete freed. Automatic test paper generation refers to questions selected from the question bank and automatically generates different kinds of papers that meet the requirements of teaching,

   so it is a typical solution process of the constraint satisfaction problem (CSP). We hve imlemented rle-bsed hierrhy whih will restrits ess t the users. The system ls delys seurity mehnisms tht rhibit dulitin f questin ers. This enbles n edutinl institute t generte questins ensuring seurity nd nn-reetitiveness f questin ers nd is bn fr rgniztins with limited stff nd resures. ur system ims t rvide fst ertins, dt strge, nd high seurity fr ll its tsks. The evlutin f trditinl nd existing Questin er Genertin systems nd the need fr n utmted system is unrveled. We hve rsed ur revised system f an Intelligent utmted Questin er Generation.

   Literature Survey is discussed in section 2. Methodology & Algorithm is available in section 3. Section 4 consists of result analysis of the project. Section 5 includes the conclusion.

2. LITERATURE SURVEYFor the automated generative system of examination papers, first, it needs to be designed by the exclusive varieties of type, the number of questions, the difficulty, and the score in order to establish the corresponding test database. Second, the papers is composed of random. Third, in a paper automatically generated, the knowledge keys concerned can now not arise. When the check paper is made, the questions are selected through the gadget within the questions database. They can meet the person’s requirements, and the performance and probability of achievement are high. The user interface of this soft is friendly. The user’s requirements can be set by way of human-computer interaction, such as: the scores for all kinds of questions in the test paper, the overall difficulty, the distribution of knowledge points and the proportion of various types of questions, and so on [9].
   1. Problem with Current Scenario:
      • Traditionally, there was no such system that would easily generate a question paper by just inserting set of question papers to the system.
      • In existing system, university use to generate question papers and distribute the question papers to respectivecolleges manually.
      • There might be chances of paper leaks due to the existing manual system.
      • The system is relatively inefficient because question papers may not reach the respective colleges on time.
   2. Limitation of Paper-based Systems: s with mst humn wrking resses, this system suffers due t bis. There might be sme questins tht re reeted in mny questin ers s the rfessr hs ersnl inlintin twrds them. S there is n gurntee f the ure rndmly generted questin er. ther rblems tht my lgue this system re the nn- vilbility f stff nd resures, nturl lmities nd idents. ls, the seurity f the system n be esily mrmised if leverge ver the ersn resnsible fr generting questin ers is btined. Other limitations include: –
      • Insufficient storage space
      • Easy to Damage
      • Inefficient document transportation
      • Supply costs
      • Poor environmental credentials
      • Limited collaboration
      • Editing problems
   3. Analysis of Paper-Based System: Frm the bve nlysis, we knw tht we need n integrted Questin er Genertin System with imrvements in seed, effiieny, ntrlled ess t the resources, rndmiztin f questins, nd seurity. In dditin, the system shuld erfrm tsks in the fstest wy withut vilting the rle-bsed hierarchy nd their ess rights liy, rvide entrl dtbse fr dt strge, ensure seurity nd timize the system’s verll performance.
3. METHODOLOGY & ALGORITHM
   1. Proposed System: T verme the existing system nmlies, this questin er genertr system is develed.
      • We resent smrt questin er generting system fr universities.
      • It is made to permit universities to generate question papers with random but even questions to cover maximum chapters of subject with difficulty level within seconds and mail thm to colleges immediately.
      • In our system we allow administrators to enter a fixed of questions and respective answers for option ticking.
      • We additionally permit admin to offer weightage & difficulty level for every questins.
      • fter this the questins re stred in dtbse lng with their weightge.
      • While generating question the admin simply has to choose the level of difficulty.
      • On this feature, the system selects questions randomly in a way that their weightage makes up for 100 marks and according to the difficulty the admin chooses the questions are chosen based on their complexity level. The questions are also added for numerous difficulty levels so that as soon admin selects the type of paper difficulty(easy, medium, difficult) the system automatically generates paper, prepares document file as per selected paper format.
      • We can also email it to other colleges. After this question paper is converted to pdf file and emailed to colleges on button click.
   2. Random Algorithm: The random Algorithm has instances to generate exam papers. One is to randomly choose questions from the question bank and then choose whether they meet the paper’s constraints. The alternative is to find out all the meet questions according to the given rules from the database, randomly selecting a number of them to constitute a take a look at the paper. Considering the low performance within the first case, in this paper, we adopt the second way to apply a random algorithm to generate test papers. In this paper, the overall parameters in the automatic test paper generation include paper title, examination time, the coefficient of difficulty, syllabus, question type and so on. Among the paper parameters, the coefficient of difficulty is one of the most important factors.

   Fig 1. Random Algorithm

   According to the analysis of the most exam results, the test score and the paper’s coefficient of difficulty are in reverse proportion. In most cases, the test scores are required to be normally distributed [5], and the average score of all trainees should be consistent with the expected score of the paper.

   Therefore, the number of questions with different coefficient of difficulty in the paper should be determined by the normal distribution function.

   The normal density function can be described as follows:

   Where x C (, +), µ and ð are constants. µ is the average of normal random variable, and ð is variance. Firstly, the expected average test score and the distribution range of scores are set according to the range of average scores, which are determined by the paper’s coefficient of difficulty, and then the average of normal random variable and variance can be figured out. Finally, the difficulty proportion of questions in the paper can be calculated according to the normal density function. So let us assume that the score range of lower and upper bounds from Easy to Difficult are li nin and li nas , i = 1,2, ,5 . And Ei are 5 probabilities followed by the coefficient of difficulty from easy to difficult.

   The variable Ei can be calculated by the function described as follows:

   To improve the success rate of automatic test paper generation, the random algorithm will appropriately adjust the parameters in the above equation when the examination questions in the question bank do not meet the requirements in the parameter table. Thus change the proportion of different difficulty of the examination questions. There are two ways to adjust the parameters. One is the adjustment of µ and ð values. The other is to adjust the values of li nin and li nas . If there is insufficient number of some coefficient of difficulty, then reduce the value of li nin properly. Otherwise, increase the value of li nas. The goal of both adjustments is to try to increase the number of questions with a larger number in question bank in the coefficient of difficulty as much as possible, and reduce the number of questions with less in question bank in the coefficient of difficulty. Thus try to avoid causing the failure of the automatic test paper generation because of the insufficient questions. The automatic test paper generation system, whose Algorithm is designed according to what is designed above, can work very well and generate test papers by setting up the parameters.

4. RESULT ANALYSISThe Automated Question paper generator has been implemented in C sharp language, which is a general- purpose, multi-paradigm programming language. The fully working system stores courses, subjects, questions andpatterns of question papers. It then applies the Algorithm to the stored question set and prints the question paper in word format. This project is implemented as a web application using Visual Studio 2019 IDE. We used Visual Studio for the Design and coding of our project. We Created and maintained all databases into SQL Server 2018, in that we created tables wrote a query to store data or records of the project.

   System Architecture of implementation is given by :

   Fig 2. System Architecture

   1. Login FormThe first web interface allows the user to select the role given while after that user needs to enter credentials to log in into the system.Fig 3. Login Form
      1. Admin: This Role has full access to the system, which includes
         • Adding Teachers
         • Adding Course & Subjects
         • Generating Question Paper
         • Directly Emailing the generated Question Paper
         • View Question Paper Logs from Master Database
         • Adding/View MCQ Question

         Fig 4: Admin Role Menu

      2. User: Teacher/Faculty is the user role here. The main objective of the user is to add the question to the database of a particular course assigned by the admin.Fig 5. Question Insertion
   2. Question Paper Generation

   After the test subjects and questions are set, the parameters of the papers need to be checked in order to ensure the correct parameter settings. After the confirmation, the user can click the button to generate the papers. If not satisfied, the system will notify the admin regarding it.

   The following process would combine the preamble information with question paper table contents to produce question paper in word format. Once all the info is passed to the system admin can move ahead and download the Question Paper.

   Fig 6: Question Paper Generation

5. CONCLUSION

In this research paper, an automated design model for Question Paper Generation has been proposed which is implemented as a real time application. The proposed work explains an automated system that shows progression from the traditional

method of paper generation to an automated process, by providing controlled access to their resources. This can be achieved by comprehension of users and their particular roles in the institute. We have considered the importance of randomization in the task of paper generation and has deployed an efficient algorithm that is completely randomized and also restricts repetition of questions in question papers. We can differentiate between administrators and subordinates by their respective tasks. Hence, the resultant automated system model for Question Paper Generation provides progression in terms of controlled access to the resources, random generation of question papers and an independent , fully secure platform.

Our system is a valuable resource for teachers in automatically generating question papers from the question repository. However, while the system designed by us stands out in all available systems, there’s scope for extra enhancements to make it more useful. For example, depending on the kind of evaluation required, the system can be made to select specific question types.

For example, if the user wants an assessment for an online quiz, it could smartly iclude all MCQs. Or, if a user is choosing

sthe term test assessment, more objective type and short answer questions should be preferred. Also, users would be overjoyed to have a feature to provide statistics for gaps in user given specifications and system- generated specs[10].Now the system is just up to generating question paper, but in the future, the system can even be implemented with separate student login for online test assessment with randomly generated questions at that moment, making it more efficient for Exam conduction.

ACKNOWLEDGEMENT

We are pleased to present the “An Intelligent Question Paper Generator using Randomized Algorithm” project and take this opportunity to express our profound gratitude to all those respective guides who guided us to complete this project. First, we express our gratitude to our project guide, Prof. Yogesh Shahare, who guided, supported and encouraged us throughout the project. We thank our college for providing us with excellent facilities that helped us to complete and present this project. We would like to thank the staff and lab assistants for helping us by giving us permission to access the computer whenever needed. We are eager and glad to express our gratitude to the Head of our I.T Dept. Prof. Swati Sinha for her support, patience and faith in our capabilities and for giving us flexibility in terms of working, reporting schedules and for her approval of this project We would like to thank our family and friends who have provided the utmost important moral support and who stood beside us throughout everything. Lastly, we would like to thank everyone who has helped us directly or indirectly in our project.

REFERENCES

[1] Vijay Krishan Purohit’, Abhijeet Kumar’, Asma Jabeen’, Saurabh Srivastava’, R H Goudar’ ,Shivanagowda, “Design of Adaptive Question Bank Development and Management

[2] Kapil Naik, Shreyas Sule, Shruti Jadhav, Surya Pandey, “Automatic Question Paper Generation System Using Randomization Algorithm”, (IJETR) ISSN: 2321 0869,Volume-2, Issue-12, December 2014

[3] W. Harlen and M. James, “Assessment and learning: differences and relationships between formative and summative assessment,” Assessment in Education, 4(3), pp.365-379, 1997.

[4] J Biggs, “Aligning teaching and assessing to course objectives, “Teaching and Learning in Higher Education: New Trends and Innovations, 2, pp.13-17, 2003.

[5] Caminero, A.C. ; Dept. de Sist. de Comun. y Control, Univ. Nac. de Educ. a Distancia, Madrid, Spain ; Hernandez, R. ; Ros, S. ; RoblesGomez, A. , “Choosing the right LMS: A performance evaluation of three open-source LMS”, Global Engineering Education Conference (EDUCON), 2013 IEEE, March 2013

[6] Ramesh, R. ; Mishra, S. ; Sasikumar, M. ; Iyer, S., “Semi- Automatic Generation of Metadata for Items in a Question Repository”, Technology for Education (T4E), 2014 IEEE Sixth International Conference, Dec 2014

[7] Liancheng Guan , “Design of the Automatic Generative System of Examination Papers Based on ARM”, International Conference on Computer Network, Electronic and Automation 2017

[8] Gauri Nalawade, and Rekha Ramesh,” Automatic Generation of Question Paper from User Entered Specifications using a Semantically Tagged Question Repository”, IEEE 8th International Conference on Technology for Education 2016

[9] Deena, G; Kothandaraman, Raja and Kaliyan, Dr. Kannan,” An Automatic Question Generation System Using Rule-Based Approach in Bloom’s Taxonomy”, Recent Advances in Computer Science and Communications 2019

[10] Krishna R; Rahila Talupula and Jahnavi Thummala, “Question Paper Generator and Result Analyzer”, Advances in Systems, Control and Automations (pp.315-323) 2021