The Study Of Forestry Yellow Page Automatic Classification

Forestry Yellow Page is one of the important forestry information resources. Forestry electronic Yellow Page not only gradually replacing the traditional paper Yellow Page, but also Contains a wealth of information. Therefore, integrating forestry electronic Yellow Page from different channels, is not only provide convenient to users to search for all types of forestry organizations information, but also lay the foundation to integrate forestry Web information and provide professional search engine. With the development of forestry information. Forestry electronic Yellow Page, especially the forestry information based on WWW, showing the characteristics of the scattered, massive, heterogeneous, non-structural diversification. Among them, text classification techniques can automatic indexing and classifying the vast amounts forestry Yellow Page, greatly improve the efficiency of information retrieval, and there is a significant value.This paper summarizes the current situation of forestry yellow Page information resources and text classification technology, summarize, analysis and research on classification of China’s forestry organizations and text features. Use improved CHI, training and learning according to data through thousands of yellow page, construct the theme characteristic thesaurus and corpus. Designed and achieve a prototype system of forestry yellow page, and classified test about2000data of forestry yellow page.The results is that the auto classification of the machine average F1value more than90%, and the classification method and system is practical. Especially for forestry organization theme feature library and corpus, and fill in the blank of research in China.