Lieff Cabraser partner Annika K. Martin has written an in-depth review of predictive coding in legal eDiscovery for Trial Magazine. In an industry where time is precious, the utilization of innovative learning technologies that are novel in the field of litigation serves as a different way for lawyers to handle their cases in a more cost-effective, accurate, and timely manner. One such type of software approach is known as predictive coding.
But what is predictive coding? It is “a type of technology-assisted review that enables a computer, relying on input from a human reviewer, to predict how documents should be classified.” Essentially, this technology calculates how the lawyer would code the remaining case documents on the basis of their input on a smaller set of documents previously reviewed. This makes discovering the relevant “hot docs” for large-data cases faster, cheaper, and more accurate as compared to traditionally manual document review.
According to the RAND Institute for Civil Justice, the cost of document review can be reduced by as much as 75% with the use of predictive coding. In addition, one study from the Richmond Journal of Law and Technology found that human reviewers missed 20-75% of all relevant documents with 90% of those mistakes caused by human error. Not only does predictive coding software serve as a more cost-effective and consistent method for document review, but it is also “more likely to follow Federal Rule of Civil Procedure 1, which encourages ‘speedy’ and ‘inexpensive’ determination of every action.”
In her Trial Magazine article, entitled “How to Stop Worrying and Love Predictive Coding,” Ms. Martin wrote, “Predictive coding cannot – and should not – entirely replace people in the search and review process. In fact, the software-training process actually requires more human interaction than other technologies, such as keyword search. While predictive coding effectively transfers the drudgery of review to the machine, the judgment remains entirely the lawyer’s. Predictive coding is better viewed as a complementary collaboration between people and machines, allowing each to do what they do best.”
Interested attorneys can read the complete Trial Magazine article here for Ms. Martin’s full exploration of the advantages and issues relating to automated document coding.