中文摘要：

为了实现Web服务请求数据的快速聚类，并提高聚类的准确率，提出一种基于增量式时间序列和最佳任务调度的Web数据聚类算法。该算法进行了Web数据在时间序列上的聚类定义，并采用增量式时间序列聚类方法。先通过数据压缩形式降低Web数据的复杂性，再进行基于服务时间相似性的时间序列数据聚类；最后针对Web集群服务的最佳服务任务调度问题，通过以服务器执行能力为标准来分配服务任务。仿真实验结果表明，相比基于网格的高维数据层次聚类算法和基于增量学习的多目标模糊聚类算法，该文的算法在聚类时间、聚类精度、服务执行成功率、聚类失真度上均获得了更好的性能。

英文摘要：

In order to achieve fast clustering of Web service request data and improve accuracy of the clustering,a Web da?ta clustering algorithm based on incremental time series and optimal task scheduling is proposed in this paper. The Web dataclustering definition in the time sequence and time series incremental clustering method are adopted in the algorithm. The com?plexity of Web data is reduced first in data compression form,and then the time series data clustering based on service time sim?ilarity is conducted. Finally,for the problem of the best service task scheduling in Web cluster services,the executive capacityof the server is taken as a standard to dispatch the service tasks. The simulation results show that in comparison with high?dimen?sional data grid?based hierarchical clustering algorithm and multi?objective fuzzy clustering algorithm based on incremental learn?ing,the algorithm proposed in this paper has obtained better results in the aspects of time clustering,clustering accuracy,suc?cess rate of all service execution and distortion degree.