← Back to Publications List

SkillMap-ML: A Data-Driven Skill Mapping and Machine Learning Framework for Classifying, Predicting, and Visualizing Digital Competency Gaps in University Students

Students & Supervisors

Student Authors
Md Siam Mehedi
Bachelor of Science in Computer Science & Engineering, FST
Jannatul Ferdous
Bachelor of Science in Computer Science & Engineering, FST
Hasibul Islam Hasib
Bachelor of Science in Computer Science & Engineering, FST
Samiha Tasnim
Bachelor of Science in Computer Science & Engineering, FST
Supervisors
Md. Mortuza Ahmmed
Associate Professor, Faculty, FST

Abstract

Popularization of digital technologies requires a great degree of digital literacy in university students to succeed in their studies and have a career opportunity. Nevertheless, there is a high level of disaggregation of digital competence that is stimulated by multi-layered socio-economic and institutional factors and requires a standardized and data-driven evaluation. To overcome this, this paper introduces the SkillMap- ML Framework, which uses a quantitative research design, based on the analysis of secondary data to categorize, forecast, and visualize the gap in digital skills. The methodology applies supervised machine learning classification (K- Nearest Neighbors and Decision Tree) to classify students into skill levels (Low, Medium, High), regression analysis to predict score of proficiency and correlation analysis to map competency structure. The Support Vector Classifier (SVC) had the highest accuracy of .95 which was very predictive as shown by the framework. Findings reveal that digital proficiency is a very predictable construct with academic engagement and technology use being good predictors. Also, correlation analysis showed that it is not a one-dimensional concept since basic and high-order scores can advance independently, which means that a digital literate person is not always an excellent one due to mere access. The results prove that resources, socioeconomic status, and institutional support are the determinants of digital readiness and present the urgent necessity of multifaceted and evidence-based interventions (such as updated curricula and equal access to infrastructures) to close the gap in structural skills.

Keywords

Digital Competence Skill Gap Machine Learn- ing (ML) Classification Regression Analysis Skill Mapping University Students.

Publication Details

  • Type of Publication:
  • Conference Name: 2nd Undergraduate Conference on Intelligent Computing and Systems (UCICS 2026)
  • Date of Conference: 17/01/2026 - 17/01/2026
  • Venue: Varendra University, Rajshahi, Bangladesh
  • Organizer: UCICS 2026
View on Publisher