Heterogeneous Educational Data Harmonization for Academic Performance Determinants Discovery in Engineering Students
In state-of-the-art benchmarks, it is noted that harmonizing heterogeneous data serves as the usual approach whereby the factors that influence the academic performance of students can be discovered. In this connection, in the case of research in Computer Science, the researcher pays attention to the use of the quantitative approach and ignores the qualitative data and the possibility of integrating both. It is also worth noting that there is an effort made to integrate qualitative and quantitative approaches in this study. The research was carried out using some sophisticated methods, including mutual information, recursive feature elimination, LASSO, PCA, Random Forest, and AdaBoost. A thorough understanding is given by combining purposefully chosen survey results, in-depth transcripts of Focus Group Discussions (FGD), and to offer a complete view. By combining qualitative themes with quantitative analytics, “heterogeneous educational data harmonization” illustrates predictive innovation and shows how it might lead to determining the main causes of students’ attainment in the early phases of their engineering education. Inimitable and harmonized methodologies were used to identify the 21 influencing determinants that were the most significant and enlightening. By adopting a unique strategy, harmonizing various data kinds, and applying the different categories of feature selection techniques, this study improved the evaluation methodologies in the field of computer science. Academicians, policymakers, and researchers will use the insights gathered from this research study.
Authors
- Dhan Maya Thapa (ORCID: https://orcid.org/0000-0002-6508-6058)
- Subarna Shakya (ORCID: https://orcid.org/0000-0003-2268-0097)
Institutions
- Tribhuvan University (NP)
- Pulchowk Campus (NP)
Publication Details
- Journal
- Contemporary Research An Interdisciplinary Academic Journal
- Published
- 2026-11-20
- DOI
- https://doi.org/10.3126/craiaj.v9i2.102350
- Primary Topic
- Online Learning and Analytics
- Type
- article
- Field-Weighted Citation Impact
- 0.00