Predictive Coding Systems for Electronic Discovery

Abstract

Over the past year, the presenters have been working to design a simple predictive coding system for electronic discovery (e-discovery) based on readily-available software and Natural Language Processing. In this presentation, they cover the history of e-discovery in the U.S., the evolution and increasing acceptance of predictive coding/Technology-Assisted Review (TAR) as part of an e-discovery process, considerations, including software choices, in constructing a predictive coding system and a demonstration of the prototype that Ms. Soundarajan developed. They will discuss their future work, which will include usability testing of the system with a focus group of lawyers who are responsible for e-discovery, the features and functionality that they would like to add to the system and the larger set of materials they would like to experiment with to further refine their system's capabilities.