Social network approach to determining privileged information in legal discovery situations

In the discovery phase of a legal action, the records of both litigants are searched for documents responsive (or relevant) to the litigation at hand and, if they are not deemed privileged as a communication providing legal advice, shared with the opposing counsel.

This is also one of the largest expense of a legal action as all of the documents must be searched and vetted for both privilege and relevance in what is often millions of documents. In this research, we build on existing approaches to e-discovery by applying the tools of social network analysis (SNA) to the problem to improve the performance of (1) responsive and (2) privilege classification engines. The industrial partner will benefit from a novel approach to solving what are two core problems of their industry and by exploring the possibility of a future social network based document analysis system.

Faculty Supervisor:

Dr. Charles Clarke


Robert Warren


Canadian Dicument Extraction





University of Waterloo



Current openings

Find the perfect opportunity to put your academic skills and knowledge into practice!

Find Projects