Thursday, June 21, 2012

Predictive coding. The future



http://ow.ly/bJkr0

A blog post by Millwright, posted on the site blog.millnet.co.uk.

This post provides a link to an audio interview taken by Chris Dale of the the e-Disclosure Information Project, which consists of an interview of James Moeskops of Millnet.  The post also provides a link to a transcript of the interview.  The interview discusses predictive coding, and the future application of technology assisted review processes in the legal profession.

The transcript quotes Mr. Moeskops as follows, "Essentially predictive coding is, as I have said, a cost and time efficient means by which to perform an initial categorisation of a collection of documents without the need to read all of the documents. Technology that attempts to categorise without the benefit of legal review input such as clustering, for instance, can be helpful within the sort of legal review exercises that are involved in regulatory investigations; however, predictive coding does go a lot further and we believe that predictive
coding can and should be used in any scenario where there are large volumes of documents to review and
categorise and in particular where the categorisation of documents is based on textual content and where
there are tactical advantages or cost advantages associated with identifying particularly relevant or hot
documents as quickly as possible; so the typical sorts of fraud investigations, price fixing investigations
and in fact DOJ and SCC investigations that have occurred and had quite high profile reviews involved in
those exercises; so I think the use of predictive coding certainly fits with the requirements of most
regulatory investigations..."

No comments:

Post a Comment