Technology Assisted Review (TAR) / Predictive Coding — New Decision

Thanks to Hon. (Ret.)  Andrew J. Peck, now Senior Counsel at DLA Piper, I just learned of a hot-off-the-presses (yesterday) Technology Assisted Review decision by the Utah federal district court.

For a quote from, and link to, the Entrata decision, check out the recent updates to the TAR / Predictive Coding Library I maintain at this blog site.

And while you’re there, you’ll see another new link — to another predictive coding online resource maintained for CTRL by Driven, Inc.’s Phil Favro.

Hi all.    Exterro has recently published what looks like a fine overview of the eDiscovery process, “The Basics of E-Discovery Guide.”  It describes a wide range of topics in easily digestible chunks.

Sprinkled throughout are quotes from various folks in the eDiscovery field.  Yours truly is one of those people.  My snippet is at the very bottom of Chapter 1, “Process“.

The Guide is also downloadable in .pdf form.

Happy reading . . . .

Check out the following items issued this week by Magistrate Judge Andrew J. Peck in Rio Tinto PLC v. Vale S.A., No. 14 Civ. 3042 (RMB) (AJP) (S.D.N.Y):

Both are discussed in this hot off the presses article by SeanDoherty in Law Technology News:  Federal Court Approves Parties’ Technology-Assisted Review Protocol, LTN (3/4/15) (“New York district court approves TAR process agreed to by parties, but acknowledges the protocol used may not apply to all TAR cases.”)

Also, in general, be sure to check out my now-updated TAR and Predictive Coding Library.

Just a quick note to remind (?or first-mind?) everyone that this site’s Resources page is an ever-expanding universe.

Some highlights of recent additions and changes include:

  • Brand new eSignatures Bibliography posted on 9/11/13
  • New/replacement Records-Retention slide deck from a webinar I did for NCC on 9/10/13
  • Revised/updated Predictive Coding & T-A-R slide deck — as revised 5/11/13

Keep visiting ITLawToday’s Resources page and the rest of this site to stay up to date on the intersection of law and IT.

Introduction to eDiscovery’s eVolution – Past, Present and Future

Three segments of the 1940 Walt Disney classic Fantasia correspond with the three historical phases business information and civil litigation discovery have undergone in the past three-quarters of a century.  Fantasia, an iconic combination of a set of Disney-animated vignettes each set to a well known classical music piece, went into production in 1938.  That same year, the first version of the U.S. Federal Rules of Civil Procedure was adopted.

In the now distant past, from 1938 until approximately 1995, business information — and thus lawsuit discovery — entailed an evolving and growing universe, mostly of paper documents.  One of my former Fenwick & West LLP colleagues,  J. Carlos Orellana, once pronounced that discovery “document review is the primordial ooze of litigation.”  Yet, even as that ooze grew and took over lawsuits, there were outer bounds.  Yet, in the next era, from 1995 to 2011 (but still the status in most circles), the explosive growth of the internet, storage chip capacities and portable devices magically rendered the scope of information and discovery vast and overwhelming.

After what seemed to some like an interminable wait, in 2012 some judges ushered in a futuristic phase.  Their respective orders addressing technology-assisted-review (TAR) – a/k/a predictive coding – were long-awaited harbingers that opposing litigators will be forced to collaborate.  Why?  To get to the merits efficiently – or at all.


I.  Distant Past – Primordial Ooze to Prehistoric Behemoths to Extinction via The Rite of Spring

Amoebae and then primitive fish were depicted throughout the beginning of the Fantasia segment The Rites of Spring, set to the music of Igor Stravinksy.  Similarly, information creation processes sputtered and slithered along for eons before eventually growing bigger and increasingly powerful.  At some point, earth’s creatures emerged from the ooze and became lumbering dinosaurs.  

Ultimately, the dinosaurs slugged it out with each other.  But to what end?  They ultimately all became extinct, as vividly displayed in the Disney Rites of Spring segment.  Similarly, as the Internet era dawned, those in the business and litigation worlds who remained wedded to paper began their own march toward obsolescence.

However large a “paper dump” ensued in a lawsuit, relatively speaking there were finite sets of discoverable materials to be collected, reviewed and produced, on the one hand, and requested, obtained and reviewed, on the other hand.  Although lawsuits lumbered along as the parties duked it out in discovery, there was more of a chance to get to the merits than nowadays.


II.  Past (& Still Present) – Taming an Army of Clones Created by The Sorcerer’s Apprentice

A non-expert, Mickey Mouse, uses his boss’ magic in the Fantasia segment Sorcerer’s Apprentice, set to the music of Paul Dukas.  The power to create goes awry as a lone broom comes to life to carry two water buckets.   Then duplication occurs and re-occurs until hundreds of identical broom-plus-buckets run amok to flood the premises.

Anologously, once paper came to be more and more displaced by electronically stored information, the size of most every company’s discoverable data set became expansive and often infinite.  Consequently, so did the vast amounts of electronically stored information (ESI) potentially subject to electronic discovery (eDiscovery) in lawsuits.  Amendments of procedural rules in the federal and states’ realms reflected were supposedly designed to shrink the scope of discovery.  But a variety of factors, including disparity in computer-savvy and financial-resources among litigation adversaries, have bogged down lawsuits more and more in a sea of data battles.

In the Disney Apprentice segment, the Sorcerer jumps in to save the day and clean up the mess.  But modern day judges have often been unable or ill-suited to readily enter the fray to tamp down on the chaos.  All judges operate under budgetary and resource restrictions; and many of them do not have sufficient technological expertise  or fierceness to ride herd on the data.

Some positive developments emerged as this next era came to an end.  For example, between 2008 and 2012, 135 judges signed onto the Sedona Conference Cooperation Proclamation (.pdf; free registration required), which urges opposing litigants and their counsel to take a collaborative approach to eDiscovery.  More significantly, a number of courts began to adopt their own orders and guidelines requiring the parties to engage in early attempts to narrow scope and cut to the chase.   For example, see:

—     the Federal Circuit’s 2011 MODEL ORDER RE: E-DISCOVERY IN PATENT CASES; and

—     the Northern District of California’s 2012 E-Discovery (ESI) Guidelines.


III.  The Future is Now –Technology-Assisted-Review Will Get the Gator and the Hippo to Join in the Dance of the Hours

A cartoon ostrich is lying down, curled up and hiding her face in the opening of the Dance of the Hours set to Almicare Ponchielli’s ballet music from the opera La Gioconda.   Eventually various other animals emerge, including a hippo who is initially terrified of an alligator.  Eventually the hippo and the alligator cooperate and even dance a pas de deux.

     

In the eDiscovery jungle, traditionally opposing counsel have been wary of each other and cooperation has been a difficult pill to swallow for many a litigator.   For many years, some tech-savvy judges and lawyers, commentators and computer scientists have been beating the drum for a better way.  For example, see:

          —     International Association for Artifical Intelligence (“AI”) and Law (“ICAIL”) (1992-2013);

         —     Collaborative Navigation of the Stormy eDiscovery Seas, 10 Rich. J.L. & Tech. 53 (2004);  and

          —     Text REtrieval Conference (“TREC”) Legal Track (2006-2013).

Between February 2012 and April 2013, half a dozen judges — in various federal and state courts nationwide — wrestled with the issue of requiring the use of technology-assisted-review (T-A-R) to help cull through one litigant’s large data set.  Five of those judges essentially issued some form of court order requiring T-A-R, a/k/a “predictive coding.  For links to all these decisions and to explanatory articles, see this T-A-R slide deck my tech colleagues and I prepared.

In sum, T-A-R involves a litigator knowledgeable about the facts of a case engaging in an iterative process to “train” a sophisticated eDiscovery platform’s AI feature to distinguish relevant documents or messages from irrelevant ones.  The promise of T-A-R is that much of the old-school document-by-document manual review by human reviewers can be avoided.  The trick is making sure the platform’s coding and algorithms work, the trainer/lawyer is effective and the people-plus-machine process is  sound as well as acceptable to the other side and the court.

While this new batch of five cases is just the start of a new era, it bodes well for the future of court-ordered cooperation aimed at forcing the alligators (plaintiffs’ counsel) and the hippos (defendants’ counsel) gracefully dancing with each other.  And when the real conductors — the clients themselves — start regularly waving their batons of efficiency and cost-effectiveness, a new futuristic cut-to-the-chase era will truly be in full swing.


Conclusion

My crystal ball shows that, in approximately a decade, eDiscovery culling and review will routinely  involve T-A-R.

In the meantime, keep returning to this new blog as we journey together to navigate ESI’s ever-changing tides.


TO LEARN MORE about T-A-R: