Noun Phrase Recognition On News Corpus

In the field of Natural Language Processing, the phrase recognition plays a vital role in machine translation, automatic index text, information extracts, text categorization and question and answer system. NP (Noun Phrase) is the major part in all kinds of phrases, therefore, NP recognition research has the strong significance and the practical value.In this paper, the current main base NP recognition method has been discussed, and then, one Chinese base NP recognition algorithm which primarily use base NP POS (part of speech) constitution template has been proposed. To identify Base NP, some linguistic knowledge is needed. The knowledge includes two parts:one is called inside knowledge, which is about the characteristic of the words that compose the Base NP; the other is called outside knowledge, which is about the context of the Base NP. These two kinds of knowledge were combined to design a system to identify Base NP. The open test on Chinese TreeBank, the public corpus, indicates the average recall and precision of 84.52 % and 94.56%.