The future is now in e-discovery: predictive coding
As people's work and personal lives become increasingly dependent on smartphones, tablets, and other emerging technology – whether to participate in the growing forms of social media or to handle day-to-day affairs – electronic discovery is more critical than ever in litigation. One of the biggest challenges in prosecuting and defending cases is how best to pursue and respond to e-discovery. The most common way that parties have attempted to identify relevant electronically stored information ("ESI") is through various types of word-searching and sampling. This approach is often woefully inadequate, typically identifying only around a quarter of the relevant documents contained within a given data source.
Therefore, the wave of the future is predictive coding. This approach identifies approximately three quarters of the relevant documents in a given data source, and that is generally equal to or better than the results of a manual review. Another important e-discovery technique is referred to as concept clustering. This is especially helpful when the responding party(ies) use code words to refer to the subject matter at issue in a case. Michael Arkfeld has a wonderfully comprehensive and practical treatise on e-discovery that discusses all of this and more: Arkfeld on Electronic Discovery and Evidence. The treatise is a must read for counsel and sophisticated litigants alike.