Electronic Discovery Question: Does Predictive Coding Case Law Require Production of Seed Sets?
In an order dated May 21, 2013 in the case Hinterberger v. Catholic Health Systems, 08-CV-380S(F), the court heard arguments from the parties about whether predictive coding case law requires the early production of a “seed set” of documents.
In this case, the parties started out sorting and culling electronic data by keyword search, but they could not agree on which keywords would suffice. At an ESI discovery status conference, the court suggested the parties use predictive coding (also called automated issue coding). The court referred to Magistrate Judge Andrew Peck’s endorsement of same in the famous case Moore v. Publicis Groupe & MSL Group, 287 F.R.D. 182 (S.D.N.Y. 2012).
In the Motion to Compel a meet and confer with defendant to establish an ESI protocol, the plaintiff asserted that Moore dictated that a seed set of documents be produced. Plaintiff contended that the defendant must provide the seed set before moving forward with predictive coding using eDiscovery software. Plaintiff argued that defendant’s refusal to do so impeded an agreed ESI protocol and the completion of ESI discovery.
However, defendant asserted that although Moore requires cooperation, the case does not specifically require a seed set be automatically produced. Further, the defendant had objections to plaintiff’s ESI consultants, and therefore, production of the seed set was premature. Plaintiff did not specifically rebut the defendant’s argument that Moore does not automatically dictates seed set production.
As the court did not disqualify plaintiff’s ESI consultants, as the plaintiff did not rebut the defendant’s reading of Moore, and as the defendant agreed to hold another meet and confer, the court denied the Motion to Compel.