Replication package for: Battling the Skills Gap: Identification of Java Knowledge Units through Machine Learning

Replication package containing:

The old and new datasets with all Java files used for training the classifiers

A Jupyter notebook for training ML-models using cross-validation

A Jupyter notebook for fine-tuning CodeBERT

The tool for scanning git repositories and extracting KUs (both front-end and back-end)

All the trained models (trained using the included notebooks)

The dataset with all the Java files and Python scripts used for evaluating the sliding window technique's performance