Technology Assisted Review (TAR) is a relatively new form of investigative litigation in which a subset of documents are reviewed by attorneys and marked for relevance. Then, specialized software is used to apply the review decisions from the subset of documents to all remaining unreviewed documents. In supervised learning TAR (SLTAR), a type of predictive coding, the software simply identifies similar documents coded as 'relevant'. In active learning TAR (ALTAR), also predictive, a subset of documents is chosen by attorneys and the computer returns samples of potential relevance. Finally, in knowledge engineering TAR (KETAR), software attempts to replicate how an attorney thinks about complex problems and then creates a statistically generated 'decision tree' using artificial intelligence algorithms to determine relevance.
According to a representative of the US DOJ Antitrust Division, they expect that use of TAR in merger and acquisition litigation will increase. Recent research has shown that TAR produces results that are at least comparable to manual review, with ALTAR and KETAR demonstrating 10 percent greater accuracy. However, such predictive coding used to determine relevancy is only as good as the algorithms used and the ability of initial reviewers. Worryingly for the scientifically minded, there is limited robust statistical data on effectiveness, accuracy, and precision.
The Knowledge Group has assembled a panel of key thought leaders to provide the audience with an in-depth review and discussion of Technology-Assisted Document Review (TAR) in Merger Investigations. Speakers will explain the benefits of adopting TAR and its implications for merging parties. The panel also will provide the audience with best practices to avoid common pitfalls and risks in using TAR for merger investigations.
To view more information or obtain a recording of this webinar, visit http://www.theknowledgegroup.org/merger-investigations/.