Analysis Of Hierarchical Structure In The Marked Compound Sentences Based On Collocation Of Relation Words

With the rapid development of information today, the Internet becomes more and more important, it has been fully integrated into the people’s study, work and life. Along with the China GDP into 10 trillion club, comprehensive national cultural soft power promotion, Chinese also play an increasingly important role in the international stage, along with big data, cloud computing era, the Internet is filled with massive Chinese data and information, so Chinese information processing has become particularly important.Chinese information processing is a branch of computational linguistics, it plays an important role in the information construction in our country. Chinese information processing is generally divided into three stages:the stage of word processing, words processing stage, sentence and text processing stage. At the present stage, our country has made breakthrough progress in the word, words processing, and are moving towards to the sentence and text processing difficult period. Compound sentence as Chinese grammar more important entity semantic information expression, it is very rich and complex, and therefore become a research focus Chinese information processing.Study on the relationship between the complex hierarchical structure, is not only beneficial to the deeper development of Chinese compound sentence information engineering, but also conducive to deepening the understanding of complex coupling law. But also to the comprehensive and accurate understanding of the complex sentence, it is necessary to clarify the hierarchy of it; once the computer to clause level structure of automatic recognition, is conducive to move forward towards the "sentence processing" stage of Chinese information processing, will also be on Chinese sentence machine translation and even text comprehension plays an important role.This paper is to analyze and do research on hierarchy and structure of the complex sentence. Firstly, a brief overview of the related concepts and theory of complex analysis, put forward the overall framework of hierarchical relationship including the relationship between word in sentence clause, annotation and analysis of structure base on collocation. Secondly, in order to grasp the clause correctly, proposed dependency parsing method, and introduces the process of analysis the relationship between word tagging rules. Finally, propose a formal model of context free grammar of the sentence, and based on this model, propose a kind of shifting and reducing algorithm based on the collocation, reducing the relationship between clauses, and generate the correspond in relation table according to the reduction, reduction can be derived from the corresponding hierarchy tree.