Abstract:

The authors presented a word disambiguation method based on local single phrase match and semantic rules via exploring the phrasestructured grammar. The method can solve the problem of redundant items existed in phrase rules and can improve the matching efficiency and disambiguation pertinence of the kinds of ambiguity by adding right nested rules between phrases and applying the matching mode of finiteautomaton. The experiment proves that the average accuracy of disambiguation reaches about 98% which is better than those of the other methods without considering the syntax and sematic between words.

Key words:  Chinese word segmentation； phrase match； ambiguous word； disambiguation method