| With the development and progress of society,information processing plays an increasingly important role in human life.After a long period of study,Tibetan information processing has been continuously studied and improved.From the most basic word processing to the application phase of machine translation and other research has been constantly forward.At present,many achievements have reached the application requirements and are put into the society to facilitate people’s use.With the progress of The Times,more and more functional applications are favored and used by people,and the market demands for Tibetan information processing technology are increasing and the requirements are becoming more and more strict.Due to the relatively late start of Tibetan information processing and the lack of researchers,it is far behind English and Chinese natural language processing.There are still some reasons,such as resources can not be shared,which results in many research results limited to closed corpus and can not be recognized by public corpus,which also hinders the development and progress of Tibetan information processing.Syntactic analysis is the study of the constraining and constrained relations among the components of a sentence.Syntactic analysis is to study the constraint relationship and constrained relationship among the components in a sentence.The final result is to represent the relationship between the components by attribute graph,which is the basis of natural language processing tasks for another tasks such as machine translation and semantic understanding.It’s depends on the result of word segmentation and part of speech tagging,and the result of syntactic analysis often affects the effect of each natural language processing task.Most of the previous studies,Tibetan nature language processing have studied by imitating English and Chinese,But Different from English and Chinese,there are abundant function words in Tibetan,which are used to connect various real words to express sentence meaning.Usually,Chinese expresses sentence meaning according to word order while Tibetan expresses sentence meaning according to function words.This paper makes full use of the features of Tibetan grammar.This paper makes full use of the functions and functions of case particles in Tibetan sentences in Tibetan grammar,and makes a comparative study of them with the Fillmore case grammar.Its characteristic is that it can fully demonstrate the grammatical rules of the Tibetan language,and it is beneficial to improve the effect of the syntactic analysis of the Tibetan language by combining statistical methods.This thesis is divided into 5 chapters.The second chapter makes a comparative study of case grammar and Fillmore case grammar in Tibetan language.Chapter three studies Tibetan word segmentation,part of speech tagging and noun phrase recognition.Among them,word segmentation is studied separately by rules and statistics,and the effect of word segmentation is compared.Part of part of speech tagging is studied by hidden Markov model,and The part of noun phrasal recognition is studied by establishing rule base。The fourth chapter studies the analysis of Tibetan sentence method through case grammar.Chapter five summarizes and prospects.Tibetan information processing has a history of more than 20 years since its inception.Looking back,from the most basic coding to the continuous progress of various applications,scholars have worked hard and made great contributions to the development of Tibetan information technology.However,due to the limitations of Tibetan grammar and the lack of various resources,the progress is slow,which requires more research and efforts.At present,many scholars have studied participles and part of speech tagging,and many results have been obtained.But Tibetan syntactic analysis is just in its infancy,and many methods need to be tried to do a lot of research. |
