<b><i>Practice Tip</i></b>:Predictive Coding and Keyword Search

Well into the era of electronic discovery, few would argue against the value of technology to assist in this regard. Predictive coding, also referred to as “technology assisted review” or TAR, is the latest attempt at taming the electronic data behemoth that presents itself as millions of pages for review (and represents approximately 70% of discovery costs). Clearly, one cannot apply the same methods or technologies that were established when a matter involved boxes of paper to electronic data volumes that are now counted in terabytes. (This article only considers the review of language-based, unstructured data and not structured data, such as financial information or documents such as schematics.)

In-house counsel are often presented with a multitude of options by both outside counsel and eDiscovery vendors, and it can be difficult to digest this information and choose the best approach for a given matter, especially when faced with the strong opinions of a trusted law firm that has successfully used a technology or process before. While outside counsel will be tasked with actually implementing the process going forward, in-house counsel can play a more involved role in the decision making process if they have a grasp of the right questions to consider. Is the use of traditional keyword searching to cull and search a corpus effectively now defunct? Is predictive coding inherently superior? What does predictive coding do differently from keyword searching and how can counsel design the most effective process around predictive coding?