Git Explorer

The problem Git Explorer solves

An opensource community is a loosely organized, ad-hoc community of contributors from all over the world who share an interest in meeting a common need, ranging from minor projects to huge developments.

One of the main hectic things in open source is to find a good repository to contribute especially issues that fall under your skillsets. It becomes more painful with beginners as they go through the process of finding projects to contribute.

A similar issue is faced by maintainers and project-owners who find it laborious to find suitable contributors whose skills align with project requirements.

A lot of time and energy can be saved if users can get personalized recommendations based on their interests and activities.

Therefore we came with this end-to-end solution for a GitHub recommendation system.

Challenges we ran into

1. Finding robust and diverse datasets in large quantities to train the ML models. We used Github API with an authorized header token to fetch more data.

2. In finding the perfect algorithm which caters to our needs. So we researched a lot and finally found Latent Semantics Indexing (LSI) technique which gives higher accuracy.

Built With

• django
• github
• google-colab
• machine-learning
• python