METHOD FOR AUTOMACTICALLY CONSTRUCTING CORPUS, METHOD AND APPARATUS FOR RECOGNIZING NAMED ENTITY USING THE SAME

摘要

Suggested are a method for automatically constructing a corpus, and a method and an apparatus for recognizing an entity name using the same, which specify unstructured personal information detection portion to have high detection accuracy, do not apply a heavy morpheme analysis method to be speedy, and construct a minimum dictionary and, if necessary, a dictionary via web search to reflect latest information. The suggested method for recognizing the entity name comprises the steps of: constructing an entity name dictionary recording the unstructured personal information; checking an entity name search result in the case that one or more among entry terms of the entity name dictionary and words inputted by a user are search targets, and extracting one or more snippets for each data characteristic; tagging the extracted snippets with corresponding entity names to secure entity name learning data; determining a learning model for recognizing the entity name, which is the unstructured personal information, based on the secured entity name learning data; and outputting a document tagged by automatically detecting the entity name, which is the unstructured personal information existing within a corresponding target document, using the entity name dictionary and the determined learning model as the target document is received.