METHOD FOR PREDICTING READING OF JAPANESE IDEOGRAM

摘要

PROBLEM TO BE SOLVED: To provide a system/method for effectively and reliably predicting the reading of Japanese ideographs. SOLUTION: A reading prediction system for operating in 'learning' and 'execution/run time' modes is provided in the implementation aspect shown as an example. In the 'learning' mode, the system operates on many input sources to generate a decision tree and uses this in the 'execution/run time' mode to return the prediction of reading with respect to an inputted Japanese sentence including Japanese ideographs. Input used in the 'learning' mode includes the reading, training corpus and quasi-phonological rules of a reference Japanese script. A basic reading and the decision tree are prepared from these inputs. In the case of operating in the 'execution/run time' mode, the system performs morphological analysis to an inputted sentence by using morphological analysis part. By using the morphological analysis, quasi-phonological rule, a basic reading and decision tree reading prediction are given.