How are we ever going to manage the enormous volumes of electronically stored information that need to be reviewed for many of today’s legal disputes? One of the most promising technologies to handle this task is “predictive coding “of electronic documents. What this means is that through the use of skilled reviewers to take random samples of the electronic documents and to code those, that the software “learns” what is relevant and what is not, so that after a statistically valid sample of the documents is coded, the computer software can very quickly and accurately replicate the choices made by the reviewers on the remainder of the database.
The results from the use of this technology are now as good as the document reviews by teams of human reviewers, while the costs are a very small fraction of the cost of human reviewers. See for example Herb Roitblat, Is Predictive Coding Defensible? This is very good news for clients facing the enormous costs of electronic document review on large cases.