Based on quotient space theory and according to page features, in the paper we first map each accessing page to pageeigenvectors, and get the content enhanced transaction set after further processing. Then we conduct the division of basic information granules on content enhanced transaction set according to its attribution, and through the citation f the collection of elementary particles, we reduce the object collection where the scanned projects set belongs while the particles are connecting and pruning in order to improve the operation efficiency of the algorithm so as to better deal with the massive data. The algorithm is suitable for the data set with small support and the high complexity.%基于商空间理论,根据页面特性,首先将每个访问页面映射为页面特性向量,进一步处理得到内容增强型Web事务集合.对内容增强型事务集合按照属性进行基本信息粒的划分,通过对基本粒子集合的引用,在粒子连接及剪枝时减少扫描项目集所在的对象集合,提高算法的运行效率,从而更好地处理海量数据,算法适用于支持度较小、复杂度较高的数据集.
展开▼
