| 摘要 |
PROBLEM TO BE SOLVED: To improve the input efficiency of a specific kind of character string by registering a part of an index as a conversion candidate obtained by KANA (Japanese syllabary)-KANJI (Chinese character) conversion, and a reading and a part of speech which are prepared in a learning dictionary in correspondence relation if the part of speech of the determined index is a specific part of speech when the index is determined. SOLUTION: For the character string of a reading inputted from an input part 11, a KANA-KANJI conversion part 14 performs KANA-KANJI conversion by referring information registered in a KANA-KANJI conversion dictionary 16 and the learning dictionary 17 to obtain an index as a conversion candidate. When a learning control part 15 determines an index as a conversion result among obtained indexes as conversion candidates, this index, and a reading and a part of speech which are prepared are registered in the learning dictionary 17 while made to correspond to each other on condition that the part of speech of the determined index is a specific part of speech. Consequently, the input efficiency of a specific kind of character string can be improved. |
