Acquiring inference rules with temporal constraints by using Japanese coordinated sentences and noun-verb co-occurrences

Abstract

This paper shows that inference rules with temporal constraints can be acquired by using verb-verb co-occurrences in Japanese coordinated sentences and verb-noun cooccurrences. For example, our unsuper-vised acquisition method could obtain the inference rule "If someone enforces a law, usually someone enacts the law at the same time as or before the enforcing of the law" since the verbs "enact" and "enforce" frequently co-occurred in coordinated sentences and the verbs also frequently co-occurred with the noun "law". We also show that the accuracy of the acquisition is improved by using the occurrence frequency of a single verb, which we assume indicates how generic the meaning of the verb is. © 2006 Association for Computational Linguistics.