| 摘要 |
<P>PROBLEM TO BE SOLVED: To provide a system and a method for analyzing a language for performing automatic analysis of right and wrong of tense of a sentence described in English by a database which stores a corpus. <P>SOLUTION: Upon extracting a sentence from a database and giving part-of-speech tags to words or phrases constituting the sentence, action verb tags showing action verbs are given to the action verbs which do not represent habitual actions among verbs, and a sorter for correlation between the action verbs and a plurality of identities regarding words to which verb tags are given is obtained. At least at an analysis part, parts of speech of the sentence for analysis described in English are presumed and its identities are converted into identity vectors. The identity vectors are given to the sorter, and the case that the verbs are the operation verb and simple present is considered as a tense error of the sentence. <P>COPYRIGHT: (C)2012,JPO&INPIT |
