Extracting decisions from discussion
Software Engineers have long, tedious discussions in email threads while working on open source projects.
These discussions tend to span over months with thousands of messages from a single user. Developers propose their appropriate ideas and speculations in these chats. The discussions contain vital project decisions which eventually play a significant role in deciding the fate of the project. Various opinions are challenged, debated or agreed upon after mindful dialogue.
Manually deciphering these long technical discussions is nearly an impossible task. This project investigates how design decisions are carried out in a development environment and ranks classified decisions according to importance. Our application automatically extracts key decisions from email discussions using machine learning techniques. Once decisions are extracted, they are analysed to predict the consequences on the fate of the project.