Please use this identifier to cite or link to this item: http://hdl.handle.net/10603/321252
Title: Recommender System for Student Learning Behaviour Based on Psychometric Data Analytics
Researcher: Burman, Iti
Guide(s): Som, Subhranil and Hossain, Syed Akhtar
Keywords: Computer Science
Computer Science Theory and Methods
Data mining
Educational tests and measurements
Engineering and Technology
Learning curve (Psychometrics)
Psychological tests
Psychology--Mathematical models
Psychometrics
University: Amity University, Noida
Completed Date: 2020
Abstract: Student s education facilitates their career and growth.Educational data mining has emerged as a new research field. To begin with, a survey has been conducted to understand the need of a system which could assist the students in enhancing their learning. The positivity of responses leads to the exhaustive review of existing models focusing on present study to analyze learning behaviour of students and to predict their grades. It then studied about various constructs affecting intellectual performance of students. The research considers psychological constructs of students related to their academic performance. To identify the causal constructs, the study undergoes exploratory factor analysis followed by confirmatory factor analysis. Based on the identified constructs, primary data is collected with the help of structured questionnaire targeting final year undergraduate students.To examine the interrelationships between the factors and to study the impact of non-intellectual constructs on student s academic performance structured equation modelling has been implemented. Different models have been constructed and the models fit values showed that students academic performance can be enhanced by revising their psychometric parameters. To test the significance of the constructs, chi square test (goodness of fit measure) is performed and meta-analysis is done using random-effects model. It is implemented to identify the average weighted correlations between academic score of a student and each nonintellectual construct. It also measures heterogeneity in effect sizes, residual deviation and standard deviation measure of effect sizes. Classifying the students, according to their academic scores, as high, average and low will help in identifying the differences in their learning behaviour. In the end the recommendation rules have been generated using decision tree (ID3) algorithm suggesting on how to improve learning behaviour of students in order to enhance their intellectual performance.
Pagination: 
URI: http://hdl.handle.net/10603/321252
Appears in Departments:Amity Institute of Information Technology

Files in This Item:
File Description SizeFormat 
01_title.pdfAttached File25.57 kBAdobe PDFView/Open
02_certificate.pdf128.26 kBAdobe PDFView/Open
03_preliminary pages.pdf191.59 kBAdobe PDFView/Open
04_chapter 1.pdf156.3 kBAdobe PDFView/Open
05_chapter 2.pdf209.86 kBAdobe PDFView/Open
06_chapter 3.pdf283.29 kBAdobe PDFView/Open
07_chapter 4.pdf1.2 MBAdobe PDFView/Open
08_chapter 5.pdf418.32 kBAdobe PDFView/Open
09_chapter 6.pdf454.78 kBAdobe PDFView/Open
10_chapter 7.pdf189.99 kBAdobe PDFView/Open
11_appendix.pdf132.78 kBAdobe PDFView/Open
12_references.pdf266.02 kBAdobe PDFView/Open
80_recommendation.pdf203.36 kBAdobe PDFView/Open
Show full item record


Items in Shodhganga are licensed under Creative Commons Licence Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0).

Altmetric Badge: