Digital documents and communications are the norm today in business. Just as companies have made this transition, lawsuits have also moved into the digital world. eDiscovery is now a common way for lawyers to gather electronically stored information (ESI) that is relevant to cases.

After the scope of an eDiscovery project is defined in a “meet and confer” session, the difficult work begins. One approach is expert manual review, where a team of people read documents and decide if information is discoverable. This process is time-consuming, expensive and inefficient

The explosion of Big Data within the business world means that huge volumes of documents are now common for eDiscovery. These large amounts of data simply can’t be reviewed without the assistance of technology. Some eDiscovery systems use keyword searches, but language by its very nature is ambiguous. As a result, keyword search results rarely include the documents that are most relevant, or relevant documents are overwhelmed by the volume of data returned in the search.

Predictive coding is a powerful tool that significantly reduces the cost and time needed for eDiscovery, while increasing the effectiveness of searches. Predictive coding relies on mathematical models to create an accurate subset of relevant documents. This eDiscovery technique has been accepted by courts, as evidenced by Da Silva Moore v. Publicis Groupe and the Global Aerospace Inc., et al v. Landow Aviation, L.P. d/b/a Dulles Jet Center cases in 2012.