A framework for data collection in educational data mining was developed with the learnings from educational data mining projects. The proposed framework discusses the educational data collection for educational data mining under three main dimensions as performance, student learning background and learning style data.
1. Performance Data: - Consists of examination marks for all the subjects over a three year period (grade 6, 7, 8). Useful in assessing weak and strong subject areas of each student. Can be used in the development of a mark prediction model.
2. Learning Background Data: - Consists of economic, sociological background, participation in assistive teaching programs and extra-curricular activities of the students. Useful in evaluating the impact of sociological background for the student performances. Can be used to evaluate the impact of assistive teaching programs for a certain subject.
3. Learning Style Data: - Evaluated based on student responses for the created questionnaire. Inspired by the LCI model. Useful in evaluating the correlation between student learning styles and their learning preferences.
Data Collection: -
About 700 student data
6 schools around the country
Targeted grade 9 students
Covered urban, suburban and rural regions
Included girls, boys and mixed schools
Created a data collection tool and conducted sessions
Research Papers Under Review - Abstract Accepted: -
1. Holistic Approach for Subject Correlation Analysis - ICDMKD (2019), (CORE Ranking : C)
2. Student Learning Preference Profiling - ICDMKD (2019), (CORE Ranking : C)
3. Student Academic Performance Prediction using Optimized Regression Based Classifiers - ICDMKD (2019), (CORE Ranking : C)
Associated Publication: -
Educational Data Mining: A Review on Data Collection Process - ICTER (2018) (CORE - Unranked, h-index: 8)
Citation: -
B. Mahanama, W. Mendis, A. Jayasooriya, V. Malaka, U. Thayasivam and T. Umashanger, "Educational Data Mining: A Review on Data Collection Process," 2018 18th International Conference on Advances in ICT for Emerging Regions (ICTer), 2018, pp. 253-258, doi: 10.1109/ICTER.2018.8615532.
