Recent Advances in Computing: Recent Advances in Natural Language Processing
CS 7301 002
Call number: 14400
TuTh 11-12:15

Instructor: Vasileios Hatzivassiloglou

This course will review recent and current research in lexical semantics, the area of computational linguistics that deals with word meaning and the properties of words that provide a basis for relating them with each other and categorizing them according to their semantic function. We will examine how lexical semantics differs from sentence-level compositional semantics and formal semantics, how lexical semantic properties control word use, substitution, and opposition, and the effect that lexical semantic properties have on word choice and implied meaning. From a computational perspective, we will look into representations of word meaning, lexical semantic components that can be automatically determined, and methods that retrieve lexical semantic knowledge from statistical evidence of the words' use in text. In the latter half of the course, we will examine applications of such automated techniques in the emerging field of bioinformatics, where computational lexical semantics techniques are used to infer the meaning and relationships between words representing genes and proteins.

Course format: Lectures interspersed with presentations by students. Students review selected papers from the primary literature and present to the class. Grading is based on the above presentations (25%), participation in the class (20%), critique of presentations by fellow students (25%), and a final written critique of a set of related papers in this area that the student chooses with the consent of the instructor (30%).

Prerequisites: A graduate course in Natural Language Processing, and familiarity with basic probability theory. Or, permission of the instructor.