摘要 |
PURPOSE:To automatically generate precise word dictionary and case frame dictionary for a supplied sentence from priority information. CONSTITUTION:An input part 11 inputs one pair of character strings from a character string storage part 10 in which plural pairs of KANA (Japanese syllabary)-KANJI(Chinese character) mixed character strings and character strings in accordance with them are housed, and a morphemic analysis part 13 finds plural KANA-KANJI mixed character string candidates by referring to the word dictionary and performing the clause disassembling and KANJI conversion of a KANA character string, and a modification analysis part 15 performs the sentence type classification of the KANA-KANJI mixed character string candidate by referring to a sentence type for the declinable word of the case frame dictionary 14, and a character string comparison part 16 compares the KANA-KANJI mixed character string candidate with an inputted KANA-KANJI mixed character string for inscription and the sentence type, and a dictionary update part 17, updates the priority information of the case frame dictionary for the word dictionary and the sentence type for coincidental inscription. |