The Metrics Behind Predictive Coding

The ESI Report’s Michele Lange, attorney and director of thought leadership at Kroll Ontrack, chats with Eric Robinson, solution architect at Kroll Ontrack, about the key metrics lawyers need to understand when using predictive coding or Technology Assisted Review (TAR) and how these metrics make e-discovery more economical and efficient.

Eric Robinson has more than 20 years of accumulated legal, e-Discovery and project management experience. As a solution architect at Kroll Ontrack, Eric works collaboratively and consultatively with clients to develop and implement strategic cost-effective, efficient and defensible discovery strategies. Leveraging his knowledge of current legal trends, regulatory matters, and information management technologies for litigation, Eric recommends defensible processes, procedures and technology solutions to optimize client efficiencies and develop best practices.

The first step to implementing predictive coding into your e-discovery review process is understanding key terms like Confidence Level, Precision, Recall, and Accuracy. And don’t worry, the intent with predictive coding is to have the mathematical values automatically computed by the document review software, no calculator required!