PURPOSE:To extract the indexes of Japanese and English sentences including new words or coined words on the basis of the small volume of a dictionary by checking the kinds of characters in an inputted bibliographic data by a dictionary part storing characters divided in respective kinds, and separating a word on the basis of the separatable condition of the word. CONSTITUTION:Respective characters grouped in respective sorts of characters such as ''kanji'' (Chinese character) and numerals are stored in a character sort dictionary part 2, respective conditions to be the divisions of words are encoded and stored in a separating condition dictionary part 4 and person's names, place names and other proper nouns are stored in an unnesessary index dictionary part 6. Bibligraphic data are inputted to an input part 1 and supplied to a character sort discrimination part 3 and a word extraction part 5. The discrimination part 3 disriminates the sort of a character while referring the dictionary part 2 and the extraction part 5 separates and extracts a word while refferring the dictionary part 4. The extraction part 5 supplies each extracted word to an index selection part 7, which selectively removes an unnecessary index in accordance with the dictionary part 6 and outputs only the necessary word as the extracted result.
展开▼
