By: J.D. Ross
(315) 443-3094

Dr. David D. Lewis

The School of Information Studies (iSchool) and the College of Law  will co-sponsor a talk on eDiscovery, to be held on October 25. The event has been organized in honor and memory of SU law professor Ted Hagelin. Hagelin passed away in May, and was an expert in technology innovation law.

Freelance computer scientist and consultant Dr. David D. Lewis will deliver a talk entitled “Machine Learning for Electronic Discovery in Legal Cases.”

Lewis will address changes in the U.S. Federal Rules of Civil Procedure and the ever-growing volume and complexity of digital data, that have led to an explosion in costs for reviewing electronic documents in legal discovery (eDiscovery).  

A recent RAND study concluded that text classification based on machine learning (so-called “predictive coding”) is the only hope for meaningful cost reductions.  Lewis will examine the history of the e-discovery crisis, what machine learning is, how it is used in discovery, and the controversy around its application.  

“Demands that machine learning be evaluated using statistical sampling procedures have led to their own controversies,” explained Lewis, “and I will briefly discuss these methods and their potential implications for legal practice.”

Lewis works in the areas of information retrieval, applied statistics, and the evaluation of complex information systems. He formerly held research positions at AT&T Labs, Bell Labs, and the University of Chicago. He has published more than 75 scientific papers and 8 patents, and is a Fellow of the American Association for the Advancement of Science.  Dr. Lewis has served as a consulting or testifying expert on e-discovery issues in civil litigation, including in the Kleen Products, Actos, da Silva Moore, FHFA, and Cambridge Place cases. 

The talk is free and open to the public. It will be held at 3:00 PM, October 25, in the Grant Auditorium in the Law School's White Hall. A reception will follow the talk, to be held in the Heritage Alumni Lounge and Rotunda.