Webinar: Building an NLP Data Factory - David Milward
David Milward is senior director, NLP technology at Linguamatics and the company’s founder. David introduced the notion of NLP and how it overcomes the problems of extracting information when faced with different names, expressions, grammars, and contexts; and then outputting the found concepts in standardized formats for use in dashboards, analysis, clustering, and feeding into machine learning (ML).
David described the steps in creating an automated NLP data factory, in a timely way and at scale, from orchestrating the disparate data input streams, through NLP processing the data, and then passing the results to dashboards, alerting systems, or another database. Linguamatics wants to get information out as fast as possible, so much is available out-of-the-box, including sources like MEDLINE, patents full text, and clinicaltrials.gov via Linguamatics OnDemand. They also provide terminologies and pre-packaged queries and query libraries for immediate use.
To address novel NLP tasks involving new questions and data sources, Linguamatics provides a comprehensive library of components and an open architecture to allow full configurability in which pluggable components can be combined to create new strategies. Capabilities include linguistic processing (built on ML and rule-based systems); terminologies; normalization; chemical name-to-structure; and table and region processing.