A Database of Relations between Predicate Argument Structures for Recognizing Textual Entailment and Contradiction

Abstract

In this paper, we present a database of logical relations between predicate argument structures (PASs) in Japanese for recognizing relations between statements. We have defined nine logical relations between PASs and manually collected argument structures and logical relations for verbs from definition sentences in a machine-readable Japanese dictionary. In addition, we augmented the relations in our database with a thesaurus of verb argument structures, which identifies synonymy and antonymy between PASs. Our database consists of 29,555 entries and 45,905 relations between PASs. In a preliminary experiment with this database, we constructed a system that recognizes synonymy between PASs in Web documents with a precision of about 0.80.