蚂蚁树在文档聚簇中的应用研究

The Application Research of Ant Tree Clustering in Documents

蚂蚁的自我聚集的行为可以形成蚂蚁聚簇,根据此行为提出一种基于蚂蚁树的文本文件的聚簇算法.算法中将对象属性作为为关键词,提取文本文件关键词组成一个关键词集合,一个集合代表一个对象(蚂蚁).算法将计算关键词的相对频率和对象之间的相似度,然后比较对象相似度阈值和相异度阈值,最终完成文本文件对象的聚簇.

The ants＇ self-aggregation behavior can form ants clustering. According to this we proposed clustering algorithm for text files. In the algorithm object attributes are the keywords, extracting keywords from text files and consisting of a set of keywords collection, a collection represents an object （ant）.The algorithm will calculate the relative frequency of keywords and the similarity between objects, compare with the dissimilarity threshold and the similarity threshold and complete the clustering of text file objects.