By Allie Philpin
Last week in New York, and sponsored by Catalyst Repository Systems, a panel discussion took place – “TAR For the Real World: Practical Problems, Pragmatic Solutions.” Hosted by Kirkland & Ellis, moderated by Monica Bay from Law Technology News, the panel featured well-known names in the legal industry – John Tredennick from Catalyst, Jason Baron from Drinker, Biddle & Reath, Conor Crowley from Crowley Law Office, Clifton Dutton from AIG Legal Operations Center, and Jeremy Pickens, a senior research scientist.
The discussion focussed principally on the way in which TAR, or Technology Assisted Review, could potentially save money for litigators. The panel also considered potential future uses for TAR that are outside the realm of e-discovery, and ways in which it can make the application of TAR much clearer and part the obstacles currently in its path.
When considering the use of TAR, the user has to accept that the process will probably review less than all the documents. John Tredennick explained: “Even if you have contract attorneys conducting 100% of the review, you still won’t have 100% accuracy.” He went on to add that e-discovery costs could be reduced by as much as 75%, allowing litigators to allocate those savings to other matters.
Much is misunderstood about the TAR process within the legal industry, which seems to be the main obstacle to its adoption. Conor Crowley believes that the primary concern seems to be that the process misses the relevant documents and litigators feel they are then put in a position where they have to re-do the e-discovery process, which can be time-consuming and expensive. But Clifton Dutton was quick to point out that the majority of attorneys already use a form of predictive coding in the guise of email system’s spam filters; the TAR process is similar.
But whereas some people think the introduction of TAR has ‘done away’ with keyword searches, this isn’t actually the case; in fact, TAR enhances keyword searches by trawling through masses of documents and allocating values to specific keywords. In an interview with Law Technology News, Tredennick said: “Our minds aren’t prepared to think about or manipulate ten thousand-word searches and figure out where to rank them. However, computers can do it well. TAR is just keyword searching on steroids!” He also added that TAR is useful for finding words similar to the keyword within documents.
What is clear is as the volume of data generated and collected by companies worldwide, the need for TAR will just grow exponentially, and it won’t be just for e-discovery purposes. The government is an obvious contender with the large volume of documents they have to deal with, along with regulatory and compliance bodies that have to accommodate increased regulation, i.e. the Dodd-Frank compliance and the Foreign Corrupt Practices Act.
At the close of the discussion, the general consensus was that TAR does have a future past e-discovery… and that it’s here to stay!