Natural Language Processing—Also Known as “Reading”
Blog: Capgemini CTO Blog
Natural Language Processing (NLP) is the holy grail of artificial intelligence, enabling computers to read and understand. But what does NLP mean for contract management and legal teams? And why are people talking about NLP in a very clinical way, as opposed to being excited?
I love language and I love English in particular. This is fortuitous because English is the only language I’m actually useful in. As a language, English is a mix of Latin, Greek, French, German and some other stuff people just made up. As such, there are multiple ways of saying almost anything. For example, the word “large” in the sentence “That is a large dog” can be substituted for dozens of words, including big, enormous, immense, huge, gigantic, etc. All have minor differences in connotation, but the meaning is essentially the same: the dog is big.
For some reason, however, people decided that simple words to describe certain things weren’t good enough and that we should use more complicated way of saying things. For example:
- “Call centers” are now “client interaction and customer relation sites”
- “Help desk” is “first level end user support”
- “Confidential information” is “non-public information held by the parties, as individuals or by employees, staff, contractors or other personnel associated with the party, blah blah blah”
Thank God these people haven’t got involved in truly mundane activities or I’d have to remind my kids to “masticate” with their mouths closed.
Why do we do this? Possibly because humans tend to think that bigger or more complicated is better. In doing so, we may have increased our collective vocabulary at the expense of sacrificing the emotional way in which we express things. Technological advancements don’t happen just because it’s fun to make new things. Technology solves problems and makes things better. So using a complicated definition of something detracts from that idea.
I now come to Natural Language Processing, or “NLP” if you want to sound cool. NLP really is the holy grail of smart machines (or artificial intelligence). Why do we call it artificial intelligence? Because it’s made by humans? Cars are made by humans but we don’t call them “artificial vehicles” to differentiate them from what nature has made, also known as “legs.”
However, I digress. According to Wikipedia (and yes, this is our new cultural canon, just accept it), NLP is “a field of computer science, artificial intelligence and computational linguistics concerned with the interactions between computers and human (natural) languages, and, in particular, concerned with programming computers to fruitfully process large natural language corpora.”
Or you could just say “computers that read and understand.”
Now that we’ve cut through the jargon, what does NLP mean in real life, for contract management and legal teams? Here are some examples from today and the near tomorrow of what NLP can do:
- Extraction – there are a number of companies on the market today that have “cognitive engines” or “brains” which can scan thousands of documents quickly and look for words, phrases or more, and put them into some kind of output. This can be used for litigation support, contract administration, spot needs in contracts (for example, looking for clauses that effect GDPR compliance, IFRS 16 clauses or whatever is exciting). These actions are currently being carried out manually by junior lawyers, contract administrators and others, but trust me, no one really wants to do it.
- Summarization or abstraction – take your extraction and get really good at it. Train the computer brain to see patterns and pull out not just key words or phrases, but the sentences that come out the most or appear the most in a piece of legislation, case law or even contract.
- Contract review – imagine your company has certain standards or a playbook that you follow. When a document comes in, you could either employ a resource to do this manually or use a cognitive engine to find the language and create your review and “redline” for you.
While NLP enables us to switch from manual effort to machine effort for high value, low-risk items, we’ll still need people to process the high-risk items. My point is that people are talking about NLP in a very clinical way, as opposed to being excited saying, “Hey, we invented robots that can read and understand—let’s do something great!” To this end, we’d love to hear more about what you’re doing with NLP technology!
To learn more about how Capgemini can inject artificial intelligence into your organization’s contracting function through NLP, contact: craig.conte@capgemini.com
Learn more about how Capgemini’s Contract Compliance & Optimization (CCO) solution provides a broader and deeper solution to your compliance, cost reduction and spend protection goals, from an often-overlooked area – the written contract.
Leave a Comment
You must be logged in to post a comment.