Maximum Entropy Text Classification Based on Key Duplication Semantic
暂无分享,去创建一个
Text classification as an important means of web search engine and text Mining was studied extensively.Most of the existing classification systems depend on the Chinese word segmentation.But with the increase in the number of web documents and continual emergence of new internet terms,the sharply increase in characteristic dimensions have serious impact on the system performance.This paper presents a new system based on the combination of key elements series of feature extraction and a maximum entropy model classification.The experiment shows that the system has better problem solving efficiency and adaptability.