| 基于K中心点的文档聚类算法 |
| |
| 引用本文： | 吴景岚 朱文兴. 基于K中心点的文档聚类算法[J]. 兰州大学学报(自然科学版), 2005, 41(5): 88-91 |
| |
| 作者姓名： | 吴景岚 朱文兴 |
| |
| 作者单位： | [1]闽江学院计算机科学系,福建福州350108 [2]福州大学计算机科学系,福建福州350002 |
| |
| 基金项目： | 国家自然科学基金，福建省自然科学基金 |
| |
| 摘 要： | K中心点算法是一个常用的聚类算法,它的主要缺陷是容易陷入局部极值,计算代价太高.本文先构造一个运用余弦相似度的K中心点文档聚类算法,然后提出一个改进算法,该算法不增加计算的复杂性,显著改进文档的聚类结果.最后,将该改进算法作为局部搜索过程嵌入到迭代局部搜索结构中,构造一个基于K中心点的迭代局部搜索文档聚类算法,进一步改进了文档聚类结果.试验结果表明该算法显著改进了文档聚类结果.
|
| 关 键 词： | K中心点算法 文档聚类 迭代局部搜索 |
| 文章编号： | 0455-2059（2005）05-0088-04 |
| 收稿时间： | 2004-06-05 |
| 修稿时间： | 2004-06-05 |
A document clustering algorithm based on K-medoid |
| |
| WU Jing-Lan, ZHU Wen-Xing. A document clustering algorithm based on K-medoid[J]. Journal of Lanzhou University(Natural Science), 2005, 41(5): 88-91 |
| |
| Authors: | WU Jing-Lan ZHU Wen-Xing |
| |
| Affiliation: | 1. Department of Computer Science, Minjiang University, Fuzhou, 350108, China; 2. Department of Computer Science, Fuzhou University, Fuzhou, 350002, China |
| |
| Abstract: | K-medoid algorithm is a popular method for clustering. Its main drawback is that it often gets trapped in local optimum and the computing cost is too high. In this paper, we first present a K- medoid document clustering algorithm with cosine similarity, followed by an improved algorithm, which can achieve the improvement dramatically without increasing the complexity of computing. Finally, we present an iterated local search document clustering alglrithm based on K-medoid, which uses the improved K-medoid document clustering algorithm as the embedded heuristic search method and gains document clustering results significantly. Experimental results highlight document clustering. the improvment achieved by our proposed algorithm in |
| |
| Keywords: | K-medoid algorithm document clustering iterated local search |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
