| 摘要 |
PROBLEM TO BE SOLVED: To increase the use efficiency of a dictionary by making it possible to output even a word, which is not present as a basic phrase conversion candidate in advance as a 1st candidate, performing conversion as a 1st candidate for a desired homophone, and reducing the capacity of a learning area while sufficiently holding the effect of learning and facilitating processing. SOLUTION: A basic phrase is determined as a 1st candidate from an inputted KANA (Japanese syllabary) character string and when learning data on a reading of an independent word is found by searching learning data according to the reading of the independent word as the base of a basic phrase, the independent word is changed according to the learning data, and when this independent word is changed into another homophone, reading data on the independent word before the change and word discrimination information on the independent word after the change are stored as learning data. Or it is judged in what state the determined independent word is used and the information on the used independent word and information on the use state of the independent word are stored as learning data. |
