基于位置关系模糊聚类的Web服务质量预测

Web service QoS prediction based on fuzzy clustering of location relation

针对Web服务质量预测受数据稀疏与噪声数据的影响,从而造成预测精度较低的问题,提出了一种基于位置关系模糊聚类的Web服务质量协同过滤预测方法。该方法利用经纬度特征通过模糊C均值聚类算法得到隶属度矩阵,采用改进的相似度计算方法分别计算用户或服务之间的位置关系相似度与历史服务质量相似度,并设置权值将两者融合,从而缓解噪声数据对服务质量预测的影响,最后使用混合协同过滤算法预测空缺的服务质量。实验结果表明在数据稀疏的情况下,该方法预测精度高于其他方法。

A collaborative filtering quality of service(QoS) prediction method based on location relationship fuzzy clustering is proposed in order to address the problem of low prediction accuracy caused by sparse and noisy data. In this method, the longitude and latitude characteristics are used to obtain the degree of membership matrix through the fuzzy C-means clustering algorithm. The improved similarity calculation method is used to calculate the similarity of location relationship and the similarity of historical service quality between users or services, and the weights are set to fuse the two, so as to alleviate the impact of noise data on service quality prediction. Finally, the hybrid collaborative filtering algorithm is used to predict the quality of service of vacancies. The experimental results show that the prediction accuracy of this method is higher than that of other methods in the case of sparse data.