| 摘要 |
PURPOSE: To display a character string in the best order for a user by storing a word which was used in the past in a learning dictionary while discriminating whether or not the word is an independent word with an attached word, and comparing a conversion candidate character sting with the information and rearranging it. CONSTITUTION: In a word storage part 4, the readings of words which are possibly independent words other than postpositional particles of Japanese and auxiliary verbs, KANJI(Chinese character) notations, and part-of speech information are stored corresponding to one another, and in a grammatical information storage part 5, group information on attached words and information on connections between words are stored. A KANA(Japanese syllabary)-KANJI conversion control part 3 retrieves words whose readings match with a KANA string read out of an input KANA string storage part 2 from a word storage part 4, and judges whether or not those words can constitute a correct output character string by using the group information on the attached words and information on the connections between the words stored in the grammatical information storage part 5. Correct output character strings are rearranged according to priority by using information in a word use information storage part 6 storing pairs of input KANA strings having section symbols interposed between independent words and attached words and KANA-KANJI-mixed output character strings. |
