© 2020 Strange Loop
With public advances such as Watson and Siri, the field of Natural Language Processing (NLP) seems poised to enter into every aspect of the way we interact with computer. But how do we make NLP flexible and practical? When people communicate online, they use their knowledge of the world to talk creatively about everything from products to news. Gaining meaning from this text means going beyond set keywords and ontologies to a deep understanding of most of the explicit and implicit ways of communicating. How can you quantify it and make decisions with it? How do you compare it and put error bars on it?
We explain how concept-based text understanding solves this problem and explain, with example and code, how you can do it yourself. We'll also explain how we've been building models of how people think about the world and how these models can be built into useful tools such as search engines, recommender systems, and classifiers.
Dr. Catherine Havasi has been researching language and learning for nearly fifteen years. In addition to being CEO of Luminoso, she is a research scientist in artificial intelligence and computational linguistics at the MIT Media Lab. She co-founded the Open Mind Common Sense project, which uses information about the world to understand natural language text and make computers easier to use. She is also active in education, previously co-founded Learning Unlimited.