Adapting Decision Trees for Learning Selectional Restrictions

This paper describes the implementation of a system that automatically learns selectional restrictions for individual senses of polysemous verbs from subject-object relationships. The selectional restrictions are inferred from an adaptation of decision tree induction, and are bound to the syntactic relations that realize them as part of a move toward automated construction of verb predicates.

• Type: Conference
• Year: 2008
• Where: FLAIRS
• Authors: Sean R. Szumlanski, Fernando Gomez
• Keywords: Artificial Intelligence, Decision Tree Induction, FLAIRS 2008, Selectional Restrictions, Subject-object Relationships