Why Deep Semantics for Natural Language Processing?

This talk will provide an overview of the program of work in semantics-based natural language processing being carried out at the Institute for Language and Information Technologies at UMBC. I will describe the Ontological Semantic text-processing environment, which seeks to create high-quality semantic representations of texts using a language-independent world model (ontology), lexicons for each language processed, a fact repository of real-world knowledge, and a suite of processors. I will motivate why I am pursuing this method of text processing despite the general tend toward stochastic methods, and will describe past and present applications for Ontological Semantic text processing.