David Han 2015 MANAGING BIG DATA IN LITIGATION DAVID HAN Associate, Morgan Lewis & Bockius, edata Practice Group MANAGING BIG DATA Data volumes always increasing New data sources Mobile Internet of Things Cloud New data types 1
EDRM MODEL MANAGING COST: MEET AND CONFER Cooperation Controlling Scope Custodians / Other Sources Time Frame Keywords / Technology MANAGING COST (CONT.) Form of Production Seek agreement up front Format may dictate process Ensure form of production is something you are comfortable receiving and creating Clawback / 502(d) orders 2
FEDERAL VS. STATE (PA) REQUIREMENTS PA ediscovery rules Order No. 564 (Pa. 2012) Discovery of ESI should be guided by proportionality and cooperation Similar in spirit to Federal Rules, but explicitly not identical to Federal Rules Amended PA Rule 4009, 4011 - Limitations as to time and scope Make ESI requests as specific as possible Nature and scope of the litigation Relevance of the ESI Cost and time burdens Relative ease of producing the ESI FEDERAL VS. STATE (PA) (CONT.) No Rule 26(f) analog - meeting prior to discovery not required No Rule 37(e) analog no Safe Harbor provision, but commentary states good faith should protect against sanctions PREDICTIVE CODING AND OTHER ASSISTIVE TECHNOLOGIES Understanding best processes and tools What is reasonable? Reasonable inquiry and reasonable steps ABA Model Rules to maintain competence, keep up to date on relevant technology 3
DEFENSE OF PROCESS Surowiec v. Capital Title Agency, Inc. (No. CV- 09-2153-PHX-DGC) (D. Ariz. May 4, 2011) Defendant used unreasonably narrow search terms Search terms must be quality control tested WHAT IS PREDICTIVE CODING? Predictive Coding machine learning Similar technologies: Spam filters Search engines Training a computer to categorize a set of data PREDICTIVE CODING CASES Da Silva Moore v. Publicis Groupe, No. 11 CIV 1279 (ALC) (AJP) (S.D.N.Y. Feb. 24, 2012) Predictive coding judicially approved, considerations: Agreement by parties to use PC Volume of ESI (3M documents) Efficacy of computer-assisted review (better than linear review or just keyword searches) Cost effectiveness and proportionality Proof of a valid process including quality control testing 4
Kleen Products LLC v. Packaging Corporation of America, No. 10 C 5711 (N.D. Ill. Aug 21, 2012) Parties disagreed as to what method to use (keyword searches vs. PC) Validating the results more important than process used Producing party is in better position to judge best method Global Aerospace Inc. v. Landow Aviation, No. 61040 (Loudoun County, VA. Cir Ct. Apr. 23, 2012) Judge allowed PC over plaintiffs objection Defendants had provided keyword testing results, showing their lack of precision, inconsistencies PC plan would include validation, 75% recall (vs. average human review recall of 59%) Proportionality, reasonable production results In Re: Actos (Pioglitazone) Products Liability Litigation, No. 6:11-md-2299 (W.D. La. July 27, 2012) Parties cooperated on PC method Both sides would review seed set and validation sets jointly 5
In Re: Biomet, No. 3:12-MD-2391 (N.D. Ind. Aug. 21, 2013) Defendant used PC after keyword filtering their collection, Plaintiff wanted re-do, PC applied to entire collection Judge ruled no, production acceptable, cost of re-work outweighed benefits Rio Tinto PLC v. Vale, S.A., et al., No. 1:14-cv- 03042-RMB-AJP (S.D.N.Y. Mar. 2, 2015) Use of TAR largely settled Transparency and cooperation still issues Validation of process can be done with statistical analysis, quality control sampling HOW TO USE PC? Technology Assisted Review Attorneys train the computer Smart prioritization of records More accurate, lower cost Potential cost shifting 6
PC WORKFLOW 1. Human Review of seed sets 2. Computer training on relevant documents 3. Iterative review of computer-scored documents 4. Statistical sampling, quality control, validation QUESTIONS? 7