Searching for Related Verbs for Semantic-based Information Retrieval

摘要

The present invention can automatically extract the related verbs used for resolving semantic ambiguity between nouns from a large amount of corpus, thereby positively excluding the subjectivity of a person, and can provide an association for semantic-based information retrieval To a verb search method.;In the field of information retrieval in recent years, Korean language processing technology has been able to detect a large number of parts, such as identification, prototype, and role in sentence, through morphological analysis and syntactic analysis. However, there is a limit to raise the accuracy and recall of retrieval to a level that a person can think with only the information of such a sentence itself. In addition, this information is used to solve semantic ambiguity between nouns in application systems of natural language processing (thesaurus, semantic network, etc.), but there are many problems in constructing such information. Until now, the information about related verbs was constructed by hand through purely human, which is a problem related to the common sense level of the builder, so the reliability of the constructed information was very low.;In order to solve the above-mentioned problem, the present invention proposes a method of automatically extracting a related verb used for eliminating semantic ambiguity between nouns from a large amount of corpus, thereby positively excluding human subjectivity, We propose a retrieval method.