Digital White Papers

May 2013: Litigation and Practice Support

publication of the International Legal Technology Association

Issue link: https://epubs.iltanet.org/i/126361

Contents of this Issue

Navigation

Page 18 of 35

TECHNOLOGY-ASSISTED REVIEW 2.0 •When you have "hiccups" in processing — for example, a document is thrown back because it's password-protected — the efficiency of TAR can decrease. Make sure processing exceptions have been addressed so you have competent text to input into the system. •Differences in time zones, character sets or formats may cause a problem while implementing technology-assisted review. When these issues arise, attempts should be made to standardize the processing before training the system. You Get What You Inspect, Not What You Expect: Once counsel is comfortable with the vendor, the process and the initial data set, he/she is ready to start the technology-assisted review process. It is important to trust the system at this point, but it is equally important to continue to verify the process is moving forward as expected. Quality control and algorithm adjustment are crucial for any technologyassisted review project. Be sure to: •Sample all levels of responsiveness. Take a random sample of documents that the system has classified as being highly likely to be responsive and a sample of those it thinks are not likely to be responsive. When feasible, one should also sample documents the system could not definitely classify. •Draw random samples across the corpus of data throughout the review project to ensure everything is staying on track. •Run keyword searches to see what comes up as responsive. Cross-reference that against the TAR results. TAR adds much-needed rigor and efficiency to the review process, arming attorneys with the tools they need to deal with the exponentially growing volume of electronic documents. When deployed correctly and surrounded by reasonable, defensible processes, technology-assisted review can deliver speed, efficiency and accuracy to a traditionally cumbersome and expensive legal process. •Perfect the algorithm. In an iterative fashion, work with the TAR vendor to adjust the system's training as appropriate. Counsel should search, test and perform quality control until he/she is comfortable the system has classified the document set effectively. Err on the side of over-inclusion rather than under-inclusion. One can also use search terms in conjunction with TAR to ensure all responsive documents are found. TAR need not be used exclusively. Ignatius Grande is a Senior Discovery Attorney and the Director of Practice Support at Hughes Hubbard & Reed LLP. He can be contacted at grande@ hugheshubbard.com. TAR TO THE RESCUE It is no longer typical for attorneys to review manually an entire body of documents in a case from beginning to end. Technology-assisted review is here to stay, and attorneys who learn to leverage mathematical algorithms, sampling and keyword search are well-positioned for success. Attorneys must also keep up with case law, which is continuously changing and setting the parameters by which TAR can and should be used. Andrew Paredes is the Senior Director of Document Review Services at Epiq Systems. He oversees the day-today operations of Epiq's document review services and assists law firms and corporate clients in the creation and implementation of review protocols, quality control and validation procedures. Contact him at aparedes@epiqsystems.com.

Articles in this issue

Links on this page

Archives of this issue

view archives of Digital White Papers - May 2013: Litigation and Practice Support