Description

This repository contains the scripts and datasets of the paper "A Text Classification Approach to API Type Resolution from incomplete Code Snippets".

Details

• The folder approach contains the data extracted that served as example in the Approach section.

• The folder datasets contains the datasets used for the external evaluation of the trained models.

• The folder discussion contains the scripts and data used in the figures of the Discussion section.

• The folder rqs groups the scripts, tools and results for each of the five research questions answered in the paper. The folders rq1, rq2, rq3 and rq5 are included here. The folder rq4 is not included since this research question is related to the benchmarking of previous research questions, and we measured the times while computing RQ2 and RQ3. In the specific case of the folder rq2 its results are not included in this GitHub repository due to size limitations. In other words, the trained models and the evaluation results are considerable large to include them here. You can access those results and models via this link: https://zenodo.org/record/7276757