In mid-November, a small group of scientists, academics, lawyers, and members of the litigation support vendor community assembled at the Gaithersburg, Maryland, campus of the National Institute of Standards and Technology (“NIST”) - a federal agency within the U.S. Department of Commerce - for the fourth year in a row.  Their mission?  To discuss the preliminary results of the 2009 TREC Legal Track, and to plan for 2010.
 
The TREC Legal Track - introduced in 2006 - is part of the larger Text REtrieval Conference sponsored by NIST.  TREC is a multi-year, multidisciplinary, international research project studying information retrieval in a variety of contexts.  The goal of the Legal Track, in particular, is to evaluate search tools and methodologies as applied to electronic discovery.  The Interactive Task of the Legal Track was designed to simulate a real e-discovery scenario in which participants use search methods of their own choosing, in consultation with a senior attorney (the “Topic Authority”), to retrieve a set of documents from a large public document collection, that they determine to be responsive to a hypothetical complaint and one or more requests for production.  This year, 11 teams participated in the Interactive Task - eight from the vendor community (including one vendor-law firm collaboration) and three academic teams.  Each team applied either proprietary or other search tools and methods to a version of the Enron data collection consisting of 847,791 documents, to see if they could exceed the results obtained through the use of Boolean key-word searches in identifying documents responsive to seven requests for production.  Final results of the 2009 TREC Legal Track will be published in or about February 2010, and will be reported here.

Why should you care about the TREC Legal Track?  Success in information retrieval is mission critical for most organizations for two reasons.  The first is knowledge management—being able to locate important records when they are needed for business purposes—and the second is being able to quickly, efficiently, and cheaply identify and produce information in response to litigation or regulatory requests.  Empirical study of search tools to learn about those that do and do not work—and under what circumstances—is the only way that lawyers and their clients can move towards automating a process that promises only to become increasingly inefficient, burdensome, and costly, as the volumes of electronically stored information (“ESI”) amass in ways that can no longer be managed through traditional, manual attorney review.

Organizations should strongly encourage their law firms and other service providers—including both litigation support software and contract review vendors—to participate in and support the TREC Legal Track.  The most promising solutions to the problems created by technology—and their attendant ESI—are likely to be technology driven.  The TREC Legal Track represents a unique opportunity to advance that agenda.  For more information about the 2010 TREC Legal Track, see http://trec.nist.gov/call2010.html or http://trec-legal.umiacs.umd.edu/.  If you would be interested in exploring how you—or your organization—can get involved in the 2010 TREC Legal Track, feel free to contact me at mrgrossman@wlrk.com.

The views expressed above are solely those of the author and should not be attributed to her firm or its clients.

About Maura R. Grossman

More ...