SYSTEM AND DEVICE FOR INTERPRETING SENTENCE BASED ON PREDICTION OF SUBJECT

摘要

PURPOSE:To analyze the content of a sentence based on prediction by designating a subject predicted from the maximum number of words in an input sentence out of predicted subjects as the subject of the input sentence and interpreting the meaning of the input sentence by referring to an effective rule in the predicted subject. CONSTITUTION:The input sentence is inputted to a subject indication word segmentation part 2 first, and is separated at every part of speech in every word, and the structure of the input sentence is analyzed according to a grammatical rule. Next, the central constituent of respective case and the word of a main verb are segmented one by one, and it is inputted to a subject selection part 3, and when it is recognized that the word is registered as a key word by referring to a subject dictionary 1, the word is taken out, and the item and the number of times of frequency of the subject are stored. And the subject with the maximum appearance frequency is outputted as the candidate of the subject of the input sentence. When the subject is decided, an interpret rule corresponding to the subject is loaded on a sentence interpretation part 5 from a rule dictionary 4, and the sentence interpretation part draws out the meaning of the input sentence by applying a loaded rule on the input sentence.