Font Size: a A A

The Chinese Statement Blocks And Disambiguation Realization

Posted on:2006-08-09Degree:MasterType:Thesis
Country:ChinaCandidate:M X ZhouFull Text:PDF
GTID:2208360152997548Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
Natural Language Processing is a cross-field subject that combined with multiple disciplines, such as linguistics, logic, physiology, psychology, computer science and mathematics etc. The aim of natural language understanding is to let the computer understand and respond human's language correctly as expected, and to build a friendly relationship between human and machine so as to realize advanced information transfer and recognition activity. With the popularity of computer and Internet, and the ongoing evolution process from traditional data and information processing to knowledge processing, more profound and comprehensive language processing techniques are increasingly required to promote the share of information and knowledge, and natural language processing has become the bottleneck of the development of society and economy. Up to now, relevant researches have only made computers understand the information of language correctly, and the intelligence of computer is still far from the level of understanding natural language as well as men. Unlike western Natural Language Processing, which was developed on an Indo-European family languages basis same as the computer was, Chinese Natural Language Processing is particularly difficult for its inherent language gap, and the feature of form-and-meaning combination and lacking of metamorphism bring more complication and obfuscation to Chinese Natural Language Processing. In order to understand a Chinese sentence, the computer must conduct syntax analyzing, semantic analyzing and pragmatic analyzing respectively, so that a formulized representation of the sentence can be produced. The process of analyzing and understanding in computer is a hierarchical process that can be divided into morphologic step, syntactic step and semantic step. The present method of syntax analysis is based on statistics, rules or the combination of them. The researches on semantic analysis based on syntax, and it focuses on developing semantic information dictionary, identifying the component in the sentences, and searching structure relations and meanings among components. These methods are widely used, but they often cause ambiguity and misunderstanding of the sentence by dividing sentence structure and semantics. This paper proposed a new method to analyze sentences and disambiguate the hierarchical structure and semantic relation. This method broke through the limitation of analyzing sentences only by syntax. It uses the theory of Three Linguistic Aspects...
Keywords/Search Tags:NLP, HowNet, knowledge base, complex feature sets, CYK algorithm
PDF Full Text Request
Related items