摘要 |
PURPOSE:To make a erroneously read character correction rapid and to make an operation simple by installing a means to replace the output sequence of plural homonymous letters corresponding to KANA (Japanese syllabary). CONSTITUTION:A character pattern inputted manually from a tablet 1 is collated by a collating part 3 through a characteristic extracting part 2 and a recognizing dictionary 4, and plural recognizing candidates for the input category are obtained. The first candidate out of them is displayed at a display part 9 as a recognizing result, and in case of the erroneous reading, the recognizing result is corrected. First, the reading is inputted from the table 1, sent to a KANA/ KANJI converting part 6, a KANJI dictionary 7 is retrieved and a KANJI converting candidate is obtained. The first candidate of them is sent to a candidate comparing part 5, compared with plural recognizing candidates and the output sequence is replaced through a sort part 8 and an erroneous character is corrected. Thus, the erroneous character can be corrected rapidly and the operation can be simply executed.
|