Bidirectional Curriculum-Industry Alignment Framework: Replication Code for a Transformer-Based Framework with Domain Disparity Measurement for Indonesian Higher Education
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
Replication archive for the manuscript “A Pilot Study of Bidirectional Curriculum–Industry Alignment in Indonesian–English Contexts Using a Domain Disparity Index” (PeerJ Computer Science, submission 137034). The study measures semantic alignment between IT curriculum learning outcomes and job-requirement sentences in both directions (Job→Curriculum and Curriculum→Job), using six sentence-embedding models and a TF-IDF baseline, and summarises the balance between the two directions with the Domain Disparity Index (DDI). Thresholds and calibrated coverage are fitted against a benchmark of 300 human-labelled sentence pairs. Contents: data/: the two cleaned corpora (423 learning outcomes from two computing programmes, labelled A and B; 776 job-requirement sentences from Indonesian IT job advertisements), with column descriptions in the README. embeddings/: normalised sentence embeddings of both corpora for the six encoders. pair_annotation/: the 300 labelled curriculum–job pairs, with the labels of annotators A and B, the third rater and the final label. language_strata/: Indonesian-dominant, code-mixed and English-dominant labels (fastText rule) for all 776 job sentences, and the 200-sentence gold sample labelled by two annotators. results/: output tables behind the manuscript, mapped table by table in the README. code/: Python scripts for the analyses and figures. make_figures.py regenerates Figures 5 to 8 from the files in the archive. figures/: all eight manuscript figures (PDF and 300 dpi PNG). notebooks/full_experiment.ipynb: the main experiment (embedding, similarity matrices, coverage). No institution, company or personal names appear in the files. The two corpora are also published on Mendeley Data (Version 3): doi:10.17632/dkrrdjs262.3 (curriculum outcomes) and doi:10.17632/zzy8v7ggkz.3 (job sentences). The README explains which input names used inside the scripts correspond to which files in this archive. Keywords
Fields of Study & Subjects
Information Systems Education and Curriculum Development
Higher Education and Employability
Topic Modeling
Publication details
- Journal / source
- Zenodo (CERN European Organization for Nuclear Research)
- Publication date
- 1 April 2026
- Language
- EN