BMI 6322 Distributional Semantics: Methods and Biomedical Applications

3 semester credit hours/meets part of advanced informatics competencies

This course concerns computational methods that learn about the meaning of words and concepts from their distribution in natural language, and consequently are able to perform cognitive tasks in a human-like manner. For example, with the appropriate learning materials, these methods have shown performances comparable with English as a second language speakers on the Test of English as a Foreign Language synonym test. Applications in the biomedical domain include information retrieval, automated indexing of the biomedical literature, literature-based knowledge discovery and the analysis of biological sequences. This course will explore the underlying theories and various methodological approaches used to measure semantic relatedness (the extent to which the meaning of two terms is related), as well as their application in biomedical and other domains. The course will provide hands-on instruction so that students will emerge with the ability to apply the methods taught in the class in their own research.