Research And Implement Of Classical Poetry Artistic Conception Classification

China has the reputation of poetry kingdom. From ancient to nowadays, myriad literati and poets had wrote many popular works, these works is the pearl of chinese literature and treasures of traditional culture. In the research of chinese poetry, the classification of poetry not only is a independent research but also is the premise and basis of content research,so the poetry classification research is a important research content in field of poetry literature.The thousands years history of china hand down myriad poetry works. In tradition research methods, researcher need read numerous knowledge material and do works of browsing、seeking、record and arrangement. Then, they need analyze and count data. Finally, they can do their judgement. The procedure not only expend time and energy but also do not get a objective result due to subjective preferences.The paper discuss the method of classical poetry classification by the theory and technology of computer and the poetry characteristic. The method can make computer understand poetry to some extent and provide reference for the research of related field. Research of the paper includes the following content:Firstly, this paper makes a summarization of computer research on the classification of poetry and introduce the concept of imago in classical poetry in detail. By the characteristic of imago in classical poetry, this paper propose the model of classical poetry artistic conception classification.Secondly, introduce the model of classical poetry artistic conception classification in detail. Because of the Tendentious of feature terms, this paper turn the feature terms into vectors that represent weight of different category. Then, the similar feature terms are clustered by the cluster algorithm and the text model is improved based on VSM and the poetry model based on feature terms clustered is formed. Futhermore, according to the poetry model, the poetry classification is converted to the text classification and is classified by the classification algorithm. The classification accuracy is contrast among different Classifier and the best Classifier is seeked. Then, the feature term based on word and feature term based on vocabulary is contrast and the result which the feature term based on word is recommended is get. Finally, experiment analyze the amount of feature term and cluster angle and define the range of their value.Finally, according to the classification, this paper build a simple classification system. After user input poetries,they can get the classification result、the best similar poetries and the feature terms data.