SYSTEM FOR EXTRACTING WORK INHERENT IN SENTENCE CORRESPONDING TO JAPANESE

摘要

PURPOSE:To rapidly and accurately extract words inherent in a sentence corresponding to Japanese by extracting the candidates of words inherent in a sentence corresponding to Japanese from a document, converging the candidates by means of language information included in the candidates, retrieving an analytical dictionary, and outputting words excluding general words. CONSTITUTION:A candidate extracting part 2 for words inherent in a sentence corresponding to Japanese extracts all original character strings corresponding to code strings as the candidates of words inherent in a sentence corresponding to Japanese and sorts the candidates extracted by the information of a changing points of character sorts in accordance with a sorting table 7. A language processing part 3 retrieves a language information table 8 and converges the candidates in accordance with the obtained information. A word selecting part 4 retrieves a Japanese analyzing dictionary 9 by means of the appearance of the candidates for the words inherent in a sentence corresponding to Japanese and selects only words not registers in the dictionary 9 as words inherent in a sentence corresponding to Japanese. A word registering part 5 writes and registers the words inherent in the sentence corresponding to Japanese in a file 6 for words inherent in a sentence corresponding to Japanese. Consequently, words inherent in a sentence corresponding to Japanese can be rapidly extracted.