Research On The Computational Methods Of Song Poetry Rules And Forms Analysis And Its Application

Computational Poetry is one of the research area of computational linguistics. The research will not only help for the field of Computer Science, but also for poetics, computational linguistics, artificial intelligence. This research work on Song Poems can be viewed as an application driver project , this paper is to do a set of research with computer assistant analysis.Our research includes the following aspects:1.Make a summarization of Computational Poetry research. And introduce the main content of prosody research. Make a summarization of research based-poetry prosody.2.Song Poems have strict limit in the rules and forms of Song Poems. According to the point, we have done some application research for it like Segmentation, Rhyme Check, Grapheme-to-Phoneme Conversion, Poetry Generation and so on.3. A solution of polyphone disambiguation which combining the modern Chinese Grapheme-to-Phoneme conversion with rules and forms of Song Poems is proposed. According to the rules and forms of Song Poems, the spell correspond to rhyme and Level and Oblique Tones. So this paper make full of use it on Grapheme-to-Phoneme Conversion. As the result shows that the correct rate is increased.4.Based on the research for subject extraction of modern Chinese text. This paper introduces two different methods of subject extraction. For the special text—Song Poems, we try the solution based on word clustering. This method analyzed the co-occurrence of words by using relativity, to create semantic relativity and generated a word cluster.As is shown by a certain number of tests, the set of research work , which combining theories and methods of computational linguistics with the field of Song Poems research, is a successful attempt. we hope that this work can serve as the foundation for further research in the field.