文本聚类算法介绍-欢迎来到重庆邮电大学理学院.doc

编 号： 审定成绩： 重庆邮电大学 毕业设计（论文） 设计（论文）题目： 学院： 学 生 姓 名 ： 专 业 ： 班 级 ： 学 号 ： 指 导 教 师 ： 答辩组 负责人 ： 填表时间：年月 重庆邮电大学教务 Propagation Clustering Algorithm (简称APC算法)、K-means算法及ad AP算法进行了深入分析。最后我们提出了基于AP 的两阶段的聚类算法，该算法既利用AP聚类快速的优点有效地减小了K-means 作用数据集的规模，又通过对初始聚类中心选取方法的优化提高了利用K-means聚类的性能。最后我们通过一系列实验将该算法与其他算法进行了性能比较和验证，最后给出了实验结论证明了我们算法的有效性。 【关键词】 中文文本聚类 AP聚类算法 两阶段 K-means 算法  ABSTRACT With the rapid development of information technology, a large number of text messages began to exist in the form of computer-readable, of which the number increased dramatically every day. How to abstract the most effective information from the vast and complex version of texts is always one of the problems to solve in information processing. How to achieve rapid and efficient information clustering, particularly the Chinese text clustering, is an important branch of the field in the information processing study of Chinese texts. Automatic text clustering technology can organize the management of text messages effectively to help people accurate positioning and efficient text messaging for users to access information needed to provide strong support. This paper introduces all aspects of text clustering in detail, mainly involving the type of text clustering, text representation, Chinese word segmentation technology, text feature extraction algorithm, the weight of the text of the characteristics of a variety of text algorithms and related technologies such as clustering algorithm. Additionally, we pay more attention to some algorithms, including Affinity Propagation Clustering Algorithm (referred to as APC algorithm), K-means algorithm, and ad AP Algorithm by analyzing the algorithms. Then we put forward our algorithm which is a two-phrase algorithm based on AP. The algorithm not only makes good use of the advantages of rapid AP clustering effectively，which reduces the size of input dat