The Samsung-Apple trial (and the recent wave of patent-infringement lawsuits in general) illustrates a key aspect of the legal profession: when it comes time to craft a viable case, a lot of attorneys will spend a great many hours reading through tons of legal documents—a process known as “discovery.” It’s painstaking and expensive work, to say the least, but also necessary.
But Big Data can help in discovery, according to an analyst David Hill of the Mesabi Group. Reading documents page by page, or searching through them via keywords, can take significant amounts of time. Predictive analytics, on the other hand, can make it much easier to mine documents (so long as they’re available in an e-format) for information.
Hill cites predictive coding as a way to successfully draw out the most useful information from all the documents present in a case. “People examine a small set of all the documents, but enough to provide a statistically reliable sample size, and classify the documents as responsive or non-responsive,” he wrote in an Aug. 8 research note. “Different predictive coding schemes exit, but the algorithms and heuristics apply their artificial intelligence, machine learning, data mining, or whatever you want to call it to classify documents that are considered responsive.”
As one would expect, the use of predictive analytics in a legal context raises some potentially thorny issues, including whether attorneys can defend the use of predictive coding tools to the court.
A survey by FTI Technology of 24 lawyers found that more than half were more likely to rely on predictive coding, in the wake of case law supporting the practice. However, more work is apparently needed to determine best practices related to its use—including whether predictive coding can meet quality control and cost standards.
“The bottom line is that predictive coding seems like a valid process technically, but its complexity suggests a deliberate, pragmatic adoption process,” Hill wrote, adding: “Although this is only one illustration of big data and its associated analytical technologies, the lesson that can be learned is this: carefully think through the drivers and inhibitors that affect the adoption of a technology, so that you can take the appropriate actions.”
Image: Andrey Burmakin/Shutterstock.com