A Syntactic Analysis Of Tibetan - Based Rules

In this paper, describing the word part of word ambiguity type and disambiguation way to do a detailed analysis, in the traditional maximum probability of matching points-based segmentation algorithm, we propose a way to use the transition probability digestion segmentation ambiguity method, also known as the transition probability disambiguation method, this method is much stronger in the disambiguation capability than the maximum probability matching. Automatic speech tagging is on the basis of well-established Tibetan HMM models, using the Viterbi algorithm. In automatic representation on the syntactic structure, the first sentence of the Tibetan language is divided into six categories, it was constructed on the basis of a scale of more than480trees library syntax tree, and then use speech string matching method to the tree represents the syntactic structure of sentences Tibetan.Taking into account the syntactic analysis in this article can not be taken as the line graph algorithms, LR algorithm and Earley algorithm that represents the syntactic structure of ambiguous sentences, it does not have the ability to analyze the syntactic structure of Tibetan ambiguous sentences, therefore, described in Chapter2,the relevant syntactic analysis when the theory of Tibetan common ambiguous sentences do classify ambiguous sentence summed up some rules exist to make syntactic analysis system in this paper have the ability to handle simple ambiguous sentences.This design parser, in accordance with the tree of speech string matching algorithm library all the syntactic structure of the sentence in the form of a corresponding tree diagram visually expressed, but because of the need to match the success of speech in order to extract the string corresponding generalized table syntax tree representation, so the parser performance and speech string matching algorithm is not much relationship, mainly have to rely on a sufficiently high accuracy rate of speech tagging system, therefore, the automatic speech tagging module design is good or bad, will directly affect the parser results...