Maura R. GrossmanSelected publications / 2011
E-discovery · Empirical study · 2011

Technology-Assisted Review in E-Discovery Can Be More Effective and More Efficient Than Exhaustive Manual Review

Maura R. Grossman and Gordon V. Cormack

Richmond Journal of Law & Technology, Vol. 17, Issue 3, Article 11 (2011)

The paper that directly tested—and rejected—the assumption that reviewing every document by hand is necessarily the most reliable way to conduct large-scale legal discovery.

Read PDFPublication recordBibTeXPlain citation
Authors
Maura R. Grossman and Gordon V. Cormack
Published
Richmond Journal of Law & Technology, Vol. 17, Issue 3, Article 11 (2011)
Format
Article / scholarly paper

Overview

Using data from the TREC 2009 Legal Track, the article compares technology-assisted review with the results that would have been obtained through exhaustive manual review by the official assessors. It shows that a process combining human judgment with automated ranking can be not only dramatically more efficient, but also more effective as measured by recall, precision, and F1.

Central contribution. At the time, automated review was often defended as a cost-saving compromise: cheaper than manual review, but presumed less trustworthy. This study changed the premise of that debate. It supplied empirical evidence that exhaustive manual review is not a gold standard and that carefully designed technology-assisted review can produce better results while requiring humans to inspect only a small fraction of the collection.

Key contributions

Abstract

E-discovery processes that use automated tools to prioritize and select documents for review are typically regarded as potential cost-savers—but inferior alternatives—to exhaustive manual review, in which a cadre of reviewers assesses every document for responsiveness to a production request, and for privilege. This Article offers evidence that such technology-assisted processes, while indeed more efficient, can also yield results superior to those of exhaustive manual review, as measured by recall and precision, as well as F1, a summary measure combining both recall and precision. The evidence derives from an analysis of data collected from the TREC 2009 Legal Track Interactive Task, and shows that, at TREC 2009, technology-assisted review processes enabled two participating teams to achieve results superior to those that could have been achieved through a manual review of the entire document collection by the official TREC assessors.

Citation

Maura R. Grossman & Gordon V. Cormack, Technology-Assisted Review in E-Discovery Can Be More Effective and More Efficient Than Exhaustive Manual Review, 17 Rich. J.L. & Tech. 11 (2011).

Related publications