个性化数据聚类的研究

Research on Personalized Data Clustering

产生无关的聚类结果,不能满足用户的实际需要,这是当前许多聚类算法存在的一个缺点。为此,该文提出根据用户实际需要进行自主聚类的个性化聚类的概念。其方法是,引入用户因素,通过设计频度向量、关心度向量来构造个性化距离函数,并与现有的著名聚类算法结合起来,给出了个性化聚类算法的一个实例描述,可以较好地解决存在的问题。

Currently,many data clus te ring algorithms usually bring out unuseful results,which can not meet users' real needs.So,a concept of personalized data clustering is put forward in t his paper.Under this con cept ,the process of data clustering is autonomous.W e introduce factor of user into clustering algorithm.And by the way of design of frequent-vector and concern-vector,personalize distance-function is const ructed.With the combination of personalize distance-function and fa mous data clustering algorithm,personalized data clustering algorithm is described,which can address the existing problems efficiently