LexisNexis, one of the largest legal research providers in the United States, is testing chatbots for lawyers, chief product officer Jamie Buckley recently told VentureBeat in a phone interview. The goal, he said, is to give users the option to take more of a conversational approach to LexisNexis research or AI services, rather than the “typing keywords into a search bar” approach that has become so common.
“Something that we’re playing with in the lab, we actually have an internal chatbot where you can start asking it questions. It replies with either an answer or what it thinks might be what you’re looking for, and it also helps you filter the results,” Buckley said. “So you might get 100,000 results on the return, but it can help to understand where are some of the differences between the results and then ask you clarifying questions based on that.”
Should LexisNexis deploy bots in this area, it will join a handful of companies also attempting to augment human lawyers with AI.
Domain-specific bots that help lawyers practicing specific kinds of law or for professionals in academia and government could also be created, Buckley said.
“So what we’re playing around with right now is what are the best legal domains for a chatbot where there are a lot of similar type questions that tend to come up and something we can really go deep in,” he said. “So in areas like that we would likely take a few specific areas and go deep as opposed to trying to go broad and try to cover everything. The chatbot won’t be very useful in that way. You have to go deep.”
Build and buy
Experiments with bots at LexisNexis are part of a larger effort by the company to marry machine learning with decades’ worth of research papers, case text, and public records — 60 billion documents in all — to provide a variety of new AI services for professionals.
To this end, LexisNexis has increased its acquisition of legal startups that use machine learning to mine documents and unearth insights.
Ravel Law was acquired in June; it focuses on legal research and doing analytics to find the best arguments to use in court. Its tools look at the language judges have cited in the past to help lawyers tailor their approach.
Intelligize, acquired in September 2016, uses machine learning and natural language processing to analyze securities law as well as SEC filings and documents.
Lex Machina was acquired in late 2015. It uses NLP and machine learning to mine through unstructured data like lawsuits and other text to get information about judges, law firms, and companies to help a law firm get clients or win a case. The company was created as part of an initiative started in 2010 at Stanford University, with support from companies like Apple and Cisco, to assist in determining the pattern of behavior of groups filing patent lawsuits.
“When LexisNexis bought us, we’d built our business around patent law, but now, using the data we have inside Lexis, we’ve expanded out to commercial law, employment law, criminal law; bankruptcy will be rolling out in a week; product liability law. So we’re going to do this really for all of law, and bring this legal analytics to all of the law,” Lex Machina CEO Josh Becker told VentureBeat.
Each company was acquired to improve NLP, machine learning, and data visualization for LexisNexis, Buckley said.
“The common theme is really helping to better enable the data-driven lawyer, so we try to synthesize, instead of just returning hundreds of thousands of results for a search. We really try to synthesize information to make it more actionable for customers, and so there’s a variety of techniques we use to do that,” he said.
News of bot tests at LexisNexis comes weeks after the launch of Lexis Answers, which suggests answers to search queries and delivers results based on past behavior. Building an AI infrastructure for the new feature took more than a year, Buckley said.