AUTOMATIC AI KANA/KANJI CONVERSION SYSTEM

摘要

PURPOSE:To enable conversion including a detailed exceptional process by preferentially displaying a character string to exceptionally be handled to an operator as a candidate of KANA(Japanese syllabary)-KANJI(Chinese character) conversion before KANA-KANJI I conversion is performed by a grammatical method. CONSTITUTION:A KANA index 'kawaga sumu' inputted by the operator is stored in an index buffer 1 and an exceptional priority learning dictionary 6 is searched for it; if no corresponding character string is found, it is temporarily stored in a nondetermination buffer 3 as a candidate, and outputted to a determination buffer 4 by an operator's judgement. When 'kawaga sumu' is not found in the exceptional priority learning dictionary 6, a KANA-KANJI dictionary 5 is searched for a candidate for a phrase and all developed in a combination candidate buffer 2. Part-of-speech information, semantics attribute information, and a syntax pattern that each word has are misconverted into 'A river lives' in this example. In this example, the input character string means 'A river is clear' and the operator corrects the converted string into 'A river is clear' and determines the string. At this time, 'A river is clear' is stored in the exceptional priority dictionary 6 for the index of 'kawaga sumu'.