Design And Implementation Of Chinese Information Processing Systems Based On Valence Theory

The Internet age brings new demands on Chinese information processing. Now,we have enough corpuses and massive data, statistical methods based on theexperience have a wide range application, but it didn’t show a particular advantagein-depth analysis. Rule-based approach finds limited in unlimited field, limitedcontrol unlimited. Words and the meaning of words are unlimited, but everyonemasters the words and meaning of words is limited; and language environment isinfinite, but language environmental factors that human use and perceive incommunication are specific.Studying Chinese information processing demands researchers have deepunderstanding of Chinese. Language is human’s expression of the awareness of things.Compared to other languages, the semantics of Chinese is more important. Fromcognitive linguistics, computational linguistics and procedures, at the height of thesemantics, this thesis use rule-based approach:(1) Classify Chinese sentence that uses Valence theory which takes into accountsboth syntax and semantics. If the language of the concrete results of the study to beable to use in computer, two requirements must be taken into account: One is toformalize and the other is general operability of the processing object. Valence theoryis an important part of cognitive grammar, can consider syntax and semantics, close tothe essence of language, can formalize and have good operability.Thought discussing the principles and methods of the Valence theory anddetermining the value of valence of commonly used verbs, it can be used to classifyChinese sentences.(2) Compare to context-free grammar which belong to formalism put forward by Chomsky in USA and phrase-based syntax put forward by Zhu Dexi which based onphase in China and comparing their grammars. It provides information forcommonality between languages and approach nature of language better. Thecomparing provide a reliable theoretical support for symbol abstracting of Chinesegrammar and establishment of production rules libraries, and provide a guarantee forthe completion of syntactic analysis(3) Establish Valence Dictionary. According to valiancy theory, this thesisestablishes Valence Dictionary and it will be used to semantic analysis(4) Focus on the three categories Chinese sentences, which include normal ordersentences, sentence with object in front, sentence with preposition "ba" and sentencewith preposition "bei"Starting from study language itself, beginning from a new perspective, this thesisis not limited in lexical analysis and syntactic analysis and parsing. Different fromtraditional good lexical and semantic analysis, it considerers the two better, and havegood inclusiveness and scalability.