模糊TOPSIS时变权重二次量化云服务推荐

FUZZY TOPSIS-BASED CLOUD SERVICE RECOMMENDATION WITH TIME-VARYING WEIGHTS FOR SECONDARY QUANTIFICATION

云服务数量的快速增长给云服务选择决策带来了挑战。为协助决策者在进行云服务选择时提供决策依据,更好地满足用户使用需求,提出一种并行模糊TOPSIS时变权重二次量化云服务推荐算法。首先,基于熵的权重赋值方法确定准则权重,解决了准则权重不好选取的问题,并采用时变权重方法作为补充设计一种新的模糊TOPSIS评估方案,时变权重的引入可以对不同时期QoS等信息的重要性进行二次量化区分,有助于提高融合决策的合理性。其次,基于时间段的模糊TOPSIS多准则融合决策信息,提出一种可并行云服务选择框架,有利于算法的实际推广应用。实验结果表明,该算法能够有效抑制不良QoS信息影响,提高服务调用成功率保证诚信服务的互操作性。

The rapid growth in numbers of cloud services brings challenges to the decision of cloud service selection,in order to help decision makers in providing decision basis for them when selecting the cloud services and to better meet the usage demand of users,we proposed a parallel fuzzy TOPSIS-based cloud service recommendation algorithm with time-varying weights for secondary quantification. First,we determined the criteria weights based on the entropy weight assigning method,which solved the problem of difficult in criteria weights selection,and the time-varying weight method was used as the complement for designing a new fuzzy TOPSIS evaluation scheme,the introduction of time-varying weight could make secondary quantified distinction on the importance of information such as QoS in different periods,this helps to raise the reasonability of fusion decision. Secondly,based on fuzzy TOPSIS multi-criterion fusion decision information of time period,we proposed a parallelisable cloud service selection framework,which was conducive to the practical promotion of the algorithm to the application. Experimental results showed that the algorithm could effectively restrain the impacts of adverse QoS information,improved the service call success rate to ensure the interoperability of faith service.