By: Daniel Gill
Machines can do almost anything these days, but they are still lacking a skill we all learn as small children: the ability to understand human language. Historically computers have looked at text merely as strings of mathematical characters, but advances in Natural Language Processing (NLP) are paving the way towards a new understanding of language. NLP enables computers to process not only the text characters themselves, but also the sentence structure, diction, and various other linguistic characteristics of human speech. NLP algorithms are applicable across many other emerging fields such as sentiment analysis, entity extraction, speech tagging, and text summarization. The wide utility of NLP means it’s one of the fastest growing markets in the technology space. Market research firms have predicted its market size may reach upwards of USD 10 Billion by the year 2020, with some estimates even higher.
The healthcare industry has some of the best use cases for NLP because healthcare companies have large troves of disparate and oftentimes unstructured databases. Traditional databases can handle well-structured medical information, but some of the most important information contained in patient records is the handwritten notes doctors scribble while treating patients. Similarly, clinical trials for drugs use structured data tables, but specific notes taken by the clinicians containing valuable contextual information are not easily ingested by a typical database. By putting NLP algorithms to work against the healthcare industry’s datasets, computers can understand and contextualize unstructured and even handwritten text, allowing end users to interact with this important data on a large scale for the first time.
The financial sector is also benefiting from advances in NLP, using its algorithms to help explain why stock prices rise and fall. Because they are often released through media outlets, leading indicators of stock fluctuations such as acquisitions, lawsuits, CEO replacements, and other newsworthy events are not easily understood by computers. But with NLP, these types of topics and conversations can be scrutinized to provide a sentiment analysis that aids investors in picking the stocks most likely to rise. And on the client side, financial institutions can use NLP to monitor social media and get a better handle on client concerns, allowing them to take care of possible issues before they become widely known.
Law firms are tapping into the power of NLP algorithms to, among other applications, expedite the process of linear review. Previously, as part of the legal discovery process, firms had to pay scores of contract attorneys to hand-review tens of thousands, and sometimes millions of documents ranging from emails to financial records in order to discover information relevant to a case. With NLP, machines can recognize and extract the semantic structure of legal documents, allowing the reviewer to hone in on relevant textual details.
Praescient has long been on the cutting edge of NLP technology and is utilizing NLP algorithms across a range of sectors to expedite the process of analysis. Working with the NexLP, a legal analysis platform that utilizes NLP and machine learning, Praescient investigated bank rate manipulation for a prominent law firm. Praescient has also used the Recorded Future social media harvesting engine which includes NLP entity extraction capabilities, to investigate ceasefire violations in Syria. And finally, Praescient hosted a countering violent extremism (CVE) live demonstration using Agolo, a text summarization tool that uses NLP to quickly distill large amounts of unstructured text articles in real time. The demo showed how analysts could sift through local media in real time to identify key CVE nodes in Kenya.
From doctor’s visits to portfolio planning, NLP is bringing groundbreaking advances to various avenues of our everyday life. Just ask Amazon’s Alexa – even she uses a form of NLP to understand your commands and help you around the house! As this field matures, Praescient will continue to scout out innovative new NLP tools and algorithms to solve our clients’ problems more efficiently.