Sense Tagging Device Based On Local Syntactic Relations and Semantic Co-occurrence Dictionary

摘要

PURPOSE: An automatic morpheme semantics tagging system is provided to automatically tag semantics of morphemes by using syntax data and semantic co-occurrence data of a declinable phrase and a noun phrase so that it enhances a meaning determination correctness. CONSTITUTION: The system comprises a short sentence recognizer(102), an auxiliary word case recovery module(103), an argument noun meaning determination module(104), and a noun phrase meaning determination module(105). The short sentence recognizer(102) recognizes a short sentence by using a basic syntax generation rule, a dependency syntax rule, an electronic morpheme meaning dictionary, a semantics frame and a declinable phrase co-occurrence dictionary generated via a morpheme analysis of an original sentence. The auxiliary word case recovery module(103) performs a case recovery for an auxiliary word by using a semantics frame and a declinable phrase semantics co-occurrence dictionary. The argument noun meaning determination module(104) determines a meaning on an argument noun. The noun phrase meaning determination module(105) determines a meaning on a noun within a noun phrase and a compound noun.