摘要： 传统语义文本分类方法难以高效地在云计算环境下进行中文文本快速分类。为解决该问题，提出一个在云计算环境下基于语义关键词提取的分布式中文文本自适应分类算法，在代理端对中文文本关键词进行基于语义的分布式提取，并将提取出的中文关键词传输到中心数据库端进行整合，从而确定文本本身类别。通过理论证明在关键词提取数量足够的情况下，提出的文本分类算法可以在保证中文文本分类效果的前提下，有效降低网络传输的代价，从而提高中文文本分类算法在云计算环境下的性能。实验结果验证了该算法在云计算环境下的可行性以及理论推导的正确性。

关键词: 云计算, 关键词提取, 文本分类, 语义计算, 网络传输代价, 中文文本语义

Abstract: Aiming to problem of the traditional text semantic classify algorithm cannot efficiently classify Chinese text under cloud computing situation, a distribution Chinese text classification algorithm based on keyword under cloud computing situation is presented, using method of distribution extracting keyword in agent-side and merging the keyword in the master database side to classify Chinese text into different kinds. Though the theoretical derivation, under enough keywords extracted situation, the algorithm can ensure the result of text classification and reduce the communication cost of network. The correctness of theoretical derivation and the algorithm is verified by the experiment.

Key words: cloud computing, keyword extraction, text classification, semantic calculation, communication cost of network, Chinese text semantic

中图分类号: 

• TP393