Research On The Syntactic Classification And Automatic Interpretation Of Chinese N+N Compound

With the development of social, science and technology, the rapid popularization of Internet and the wide application of new media, new things continue to emerge and merge, Chinese Noun+Noun compounds is also increasing. Followed by this, the study of Chinese Noun+Noun compounds is also gradually popular, Chinese Noun+ Noun is an essential issue in many Natural Language Processing (NLP) tasks, such as Information Retrieval, Machine Translation, Question Answering systems and so on. In details, the main contents of this paper include the following aspects:1. The syntactic classification of Noun+Noun Compound based on rules. The four basic relations between compounds are modifier-head relation, coordinate relation, appositive relation and subject-predicate relation. The purpose of syntactic classification of Noun+Noun Compounds is to determine the syntactic relations of the Compound. This paper summarizes the combination rules of different semantic relations of Noun+Noun compounds, makes use of HowNet, SKCC and Cilin, from the perspective of computational linguistics, we first realized the syntactic classification of Noun+Noun Compounds based on the statistics of the rules.2. Automatic interpretation of Noun+Noun Compound based on paraphrasing verbs. Noun compound automatic interpretation is to create the paraphrases by recovering the implied verbs between the nouns and using the paraphrasing templates. We use the approach based on the paraphrasing verbs, Xinhua corpus from 1991 to 2004 and the HowNet to acquire the paraphrase of Chinese Noun+Noun compounds, and filter them through the Web.3. The design and implementation of syntactic classification and automatic interpretation of Chinese Noun+Noun. We initially realized the system of syntactic classification and automatic interpretation of Chinese Noun+Noun compounds and hope the result of the paper can be applied in some other fields such as information retrieval, machine translation and soon.4. The application of the research of Noun+Noun Compound in dependency parsing and machine translation. The first part utilize the results of syntax classification of Chinese Noun+Noun compounds, we extract sentences from Xinhua corpus from 1991 to 2004 and parse them by the dependency parser. Through analyzing the dependency errors of Noun+Noun in dependency parsing, we can find out the deficiency in parsing Noun+Noun compounds of the dependency parser and finally correct the parsing errors. The second part utilize the results of Automatic interpretation of Noun+Noun compounds, we translate the Chinese Noun+Noun compounds, the paraphrase of Noun+Noun compounds and the sentences which N+N replaced by interpretation phrases by using the Google and Baidu online translation system. Through evaluating and scoring the results of translation, we verify the validity of the research of automatic interpretation of Noun+Noun compounds in machine translation.