摘要
Web用户聚类是实现自适应网站和为用户提供个性化信息服务的关键技术之一。将FCM算法与差分进化算法相结合,提出一种用于解决web用户聚类问题的混合算法,改进了适应度函数,并采用局部搜索策略进一步增强算法的寻优能力,加入FCM优化操作加速算法的收敛。实验表明,该算法全局搜索能力强,实现了较好的用户聚类效果。
Web user clustering is one of the most important technologies to develop the adaptive site and to provide individualized information service.A new hybrid algorithm based on differential evolution and fuzzy c-mean algorithm is proposed to solve the problem of web user clustering.In the new algorithm,the fitness function is improved and the local search strategy is applied to enhance the ability of finding optimal solution.In order to accelerate the convergence,the FCM optimizing operation is utilized.Experimental results reveal that the proposed algorithm has strong capability of global search and can get the satisfactory performance in user clustering.
出处
《电脑知识与技术》
2011年第10X期7452-7454,7456,共4页
Computer Knowledge and Technology
