awesome-machine-learning-on-source-code

Why we included this project

Machine learning on source code is a niche worth knowing about if you build developer tools, and this list is the handiest single map of it. It collects research papers, datasets, talks, and software projects, organized by sub-topic, so you can jump straight to code completion, program repair, clone detection, or program synthesis without wading through unrelated work. The list was maintained by source{d}, a company that built ML tooling for code, and the selections show it: they reflect real engineering problems rather than academic fashion. It is no longer updated, and the maintainers point to ml4code.github.io as the active alternative, but the papers and datasets gathered here still give a solid grounding in the field before you start building your own code-analysis pipeline.