Scholarly Commons @ FAMU Law

Article Title

The Past, Present, and Future of Predictive Coding

Authors

Matthew G. Kenney

Abstract

Electronic discovery, or e-discovery, refers to the discovery of electronically stored documents and images.' Examples of e-discovery related documentation would include email, digital versions of paper documents (e.g. MS Word, PDF, Excel, and PowerPoint), social media postings, digital photos, Global Positioning System data, and content within computerized databases, etc. Digital data stored on computers, smartphones, tape drives, hard-drives, portable digital storage devices and the like would fall under the domain of e-discovery. Collecting and sorting massive amounts of electronically stored data presents both opportunities and challenges for lawyers.

For context: In 2015, electronic discovery was a $10.2 billion global industry. Of this amount, $8.2 billion flowed to e-discovery service providers (e.g. document review by contract attorneys and vendors); and $2 billion flowed to the development of new software. The worldwide e-discovery market is expected to grow at 9.4% annually through 2019.

Predictive coding-a dimension of e-discovery-is a process whereby attorneys train computer programs to identify potentially relevant documents within a large body of documents.This process begins with the attorney(s) selecting a "seed set" of documents and choosing keywords relevant to the case. This seed set is then searched, and re-searched, in an iterative process until the software recognizes patterns. For practical purposes, it is simply telling the software what to find. However, attorneys need to have a general understanding of statistics and computer assisted technology prior to engaging in predictive coding.

Recommended Citation

Matthew G. Kenney, The Past, Present, and Future of Predictive Coding, 12 Fla. A&M U. L. Rev. 165 (2016).
Available at: https://commons.law.famu.edu/famulawreview/vol12/iss1/7