unsupervised-sentiment-analysis-on-GitHub-data

In this project, I performed sentiment analysis using an unsupervised machine learning approach on data collected from the version control website GitHub. I explored the distributions and relationships between projects commit comments sentiments, and project features like watchers, comments time of day and the repository programming language. Data wrangling steps are performed to prepare the data for analysis. Sentiment score calculations for each commit comment is calculated, in which it required a number of preprocessing and score estimation steps.