This repository consist of two main projects. 1- GithupAPi Fetcher 2- Github Success Predictor.
The Fetcher was constructed to use the api to gather data about multiple projects on github. We choose random projects based on their stars and pull requests. 1000 project's meta data for five languages where fetched and organized in a dataset file. This dataset file is freely available for research purpose here -> http://junaidmaqsood.com/success-or-failure-identification-for-githubs-open-source-projects
Then the dataset was used to create and test on mltiple machine learning algorithms. The best model with best recal and accuracy was selected and feed to the Github Success Predictor Project. This project uses the WEKA library and the model generated by the dataset we coleected to create a prediction model of github. This is able to take your repository meta deta and classify that into either successfull or un-successfull project. All these classification are ML Based and represent the latest trends.
Future Update: 1- Instead of filling in all the parameters, The software takes a link to your repository and fetches your projects meta deta and uses that to classify. Restriction (Will require internet connection)