This course will explore methods for extracting different types of information from natural language text. We will study algorithms to extract semantic concepts (e.g., named entities, hypernym/hyponym classes, semantic roles), facts (e.g., event roles, event chains, factual relations), and opinions (e.g., sources, targets, aspects). The course will include methods for extracting information from text corpora as well as the Web, and will emphasize applications for both broad-coverage and domain-specific information extraction. Specific topics will include:

• Named Entity Recognition
• Information Extraction from the Web
• Relation Extraction
• Semantic Role Labeling
• Semantic Lexicon Induction / Set Expansion
• Semantic Taxonomy Induction
• Event Extraction
• Narrative Event Chains
• Opinion Source and Target Extraction

This course will presume knowledge of basic natural language processing (NLP) concepts. Consequently, the NLP course (CS-5340/6340) is listed as a prerequisite. However, anyone who has familiarity with NLP is welcome to take the class! If you haven't already taken CS-5340/6340 but want to take this class, please email the instructor.