Analyzing the gap between its curriculum and industry demand in Nepal using NLP

Abstract

Bridging the gap between academic IT curricula and industry requirements remains a pressing challenge in Nepal, given the rapid evolution of technology. This study presents a systematic, data-driven approach to evaluate the alignment between IT programs and industry needs using Natural Language Processing (NLP) techniques. Course descriptions from leading universities were analyzed alongside IT job postings using BERT-based embeddings to calculate semantic similarity, enabling the classification of courses into high, medium, or low alignment categories. Concurrently, KeyBERT was employed to extract key skills, tools, and technologies from both datasets, allowing for an assessment of keyword coverage. By integrating semantic similarity scores with keyword overlap, a comprehensive alignment metric was developed, capturing knowledge domains, technical skills, and professional competencies.

The analysis identified measurable gaps, including a knowledge Gap of 0.40, a Technical Skills Gap of 0.25, and a Competency Gap of 0.40, yielding an Overall Gap of 0.34 and an Alignment Index of 0.66. These findings indicate moderate alignment while highlighting deficiencies in emerging technologies and employability skills. The methodology offers a reproducible and quantitative framework for curriculum evaluation, providing actionable insights for educators, curriculum developers, and policymakers to enhance program relevance and better prepare graduates for the evolving demands of the IT industry.