Search |
Interesting information |
The Limitations of Keyword Searching
Bruce N. Furukawa, the technology partner at Severson & Werson and a member of the firm's Litigation Practice Group, notes that: "There is no question that keyword searches can help identify potentially relevant data within a data set. However, more frequently, cases involve such large volumes of data that relying on keyword searching alone is a flawed practice. We went with a solution that came equipped with text analytics, enabling features such as concept search, clustering, and similar document detection. In particular, we used the similar document detection feature to directly pinpoint a number of hot documents in the opposing party's production that keyword searching never would have found.
"Since this matter, we have continued to use text-analytics features to inform our litigation and e-discovery strategies. Features such as concept clustering and categorization have proven to be invaluable. Clustering allows us to automatically sort documents into conceptually related subsets. Categorization organizes the documents based on a small set of examples that our users identify."
(Please see: "Concept Searching Saves Clients Thousands," by Bruce N. Furukawa, at LAW.COM, 2 April 2010)