Analysis Of Sentence Pattern Of Simple Declarative Sentence In Modern Mongolian Language Based On Corpus

The paper selected all parts of simple declarative sentences of language at primary schools in corpus of modern Mongolian language of one million words (990 sentences) and carried out the sentence analysis, regarding the research of traditional Mongolian grammar and regarding Modern language information processing method as a technical reference. It concentrated on observing structure of sentence, type of sentence patterns, frequency of use of a variety of sentence patterns, constructed the library of sentence pattern of simple declarative sentence in modern Mongolian language. The library contains 134 kinds of sentences. Each sentence pattern includes the boundaries of Sentence Elements, the order of sentence elements, the level of Sentence Elements, the constituent elements of the sentence elements, sentence patterns, sentence cases. Based on the above-mentioned analysis results of sentence elements in 990 sentences, it summed up the Judge rules of subject, predicate, attribute, object, adverbial respectively and established sets of Recognition rules of sentence elements of simple declarative sentence in modern Mongolian language. Among them, there are 72 recognition rules of subject,94 identification rules of the predicate,73 identification rules of object,50 attributive identification rules,69 adverbial identification rules,26 recognition rules of non-sentence elements. Based on the set of rules, the paper developed automatic annotation Program of sentence elements of simple declarative sentence in modern Mongolian language and Tried to do automatic annotation of sentence elements (6300 sentences) in a part of simple declarative sentence on corpus in modern Mongolian language of one million word and reached an initial accuracy rate 85.6%.Full-text contains Chapter 5.Chapter 1, Introduction, it briefly described research significance, research base, research methods of sentence patterns of simple declarative sentence in modern Mongolian language.Chapter 2, it briefly described the theoretical background issues.on the Mongolian sentences, simple sentences, declarative sentences, sentence patterns and so on.Chapter 3, it described the Methods analysis of sentence and form of corpus of sentence patterns of simple declarative sentence in modern Mongolian language.Chapter 4, it introduced the construction process and the actual contents in Library of sentence patterns of simple declarative sentence in modern Mongolian language.Chapter 5, it described the process which it established sets of identification of sentence elements of simple declarative sentence in modern Mongolian language and situations which it developed automatic annotation process of sentence elements of the simple declarative sentence in modern Mongolian language.With auto-tagging attempts, it explained the focus and difficult problems which it need be concerned about Sentence analysis research on information processing.It summarized the text and presented the work prospects in study of sentence patterns in modern Mongolian language.In summary, this study is a basic research area in Mongolian information processing. The library of sentence patterns of simple declarative sentence in modern Mongolian and automatic dimensioning procedure of sentence elements are indispensable knowledge resources and tools on Mongolian information-processing infrastructure. From a macro point of view, analyzing sentence, tagging sentence elements, confirming the sentence pattern are only way to applications links of Realization of machine translation, natural language understanding; From the microscopic point of view, the knowledge resources which they provided help to check with the relationship between sentence elements, determine the structural integrity of sentence.