Natural language processing (NLP) is among the fastest growing applications of artificial intelligence (AI). It’s also one of the most difficult to build. NLP powers a growing number of tools, such as chatbots, virtual assistants like Amazon’s Alexa, and even the spell check for the communication apps we use to send text messages from our devices.
But it’s fraught with challenges because to create a machine learning (ML) algorithm requires you to “teach” a machine how to interpret, distill and generate language almost spontaneously. Compound that challenge by what’s well known among NLP developers: software cannot pick up on subtlety. In our work, we know data cleaning for NLP requires context. For example, if workers are labeling your text data, you would want them to tag the word “bass” accurately, knowing whether the text in a document relates to fish or music.
Legal is among the many industries that are looking for ways to apply AI to create operational efficiencies. And, of course, most legal documents are text. Now consider this: a machine can review in seconds what would take an expensive attorney hours to review. Heretik, a company based in Chicago, Ill., uses machine learning (ML), specifically NLP, to automate tasks related to the legal domain, such as contract review and project forecasting.