ML-powered 🤖 for finding and labeling good first issues in your GitHub project!
GFI-Bot
A GFI-Bot introduction paper is available as follows (submitted to ESEC/FSE 2022 Demonstration Track):
- Hao He, Haonan Su, Wenxin Xiao, Runzhi He, and Minghui Zhou. 2022. GFI-Bot: Automated Good First Issue Recommendation on GitHub. Currently Under Review at the ESEC/FSE 2022 Demonstration Track. https://hehao98.github.io/files/2022-gfibot.pdf
The underlying ML approach is introduced in the following paper:
- Wenxin Xiao, Hao He, Weiwei Xu, Xin Tan, Jinhao Dong, and Minghui Zhou. 2022. Recommending Good First Issues in GitHub OSS Projects. In Proceedings of the 44th International Conference on Software Engineering, ICSE 2022, Pittsburgh, PA, USA, May 21–29, 2022. ACM. https://hehao98.github.io/files/2022-recgfi.pdf
We provide a good first issue recommendation dataset at Zenodo.
Get Started
GFI-Bot is available at https://gfibot.io, where you can browse through existing good first issue recommendations or register your own repository for recommendation. GFI-Bot can be installed in GitHub repositories from the GitHub App page.
NOTE: GFI-Bot is currently in pre-alpha stage. It is undergoing rapid development and still highly unstable. We cannot guanrantee the preseveration of registered users and repositories in the next release and it may have unexpected behaviors on GitHub. We will change this note after GFI-Bot reaches a certain level of maturity
Roadmap
We describe our envisioned use cases for GFI-Bot in this documentation.
Currently, we are focusing on the following tasks:
- Identifying an optimal training strategy
- Improving user experience
Development
More development details are available at https://github.com/osslab-pku/gfi-bot