Project: Splitting Task Oriented Social Networks into a Task Related Layer and the Rest
Mentor: Vladimir Filkov (Computer Science)
Task Oriented Social Networks are distributed groups of people who participate together on accomplishing specific tasks (e.g. Open Source Software projects), linked in pairs based on their communication. The links in these networks are multiplexed: some communication is about specific tasks while others are about the process of accomplishing them or of a more social nature. We have assembled historically complete social networks of communications between developers in hundreds of Open Source Software projects. The goal of this RTG project is to develop methods to split each of those networks into two layers: links used for communicating task information, and the rest.
The student on this project will do some basic content analysis of the text messages to identify task-oriented keywords, and will develop the appropriate data structures and analysis framework to store and specify longitudinal multiplex networks. The initial work and results along these lines is given in our paper "Latent Social Structure in Open Source Projects", http://macbeth.cs.ucdavis.edu/cathbaz.pdf.