Predictive coding describes a review tool that employs algorithms to determine the relevance of documents based on the “training” the computer receives from human reviewers.
“Based upon the coding of a sample or seed set of documents by a team of lawyers with subject-matter expertise, the computer learns what is considered responsive versus nonresponsive and then applies that logic to the remaining documents to suggest additional documents that are like the first responsive set,” says Jacquelyn Caridad, of counsel to Morgan, Lewis & Bockius.
