基于用户需求特征的Web服务动态组合方法研究  

Research of web services combination recommendation based on user demand characteristics

作　　者：袁士君 艾中良[1] 李喻 

机构地区：[1]华北计算技术研究所,北京100083 [2]北京航空航天大学,北京100191

出　　处：《软件》2015年第3期69-74,共6页Software

摘　　要：为了解决以满足用户需求为目标的动态服务组合问题,适应Web服务组合过程中的动态性、不稳定性,提出了一种基于用户需求的改进蚁群算法。算法包括两个优化的过程,局部优化规则和全局优化规则,分别保证得到局部的最优解和全局的最优解,保证服务选择过程的趋优性,尽可能满足用户对服务的多个需求特征要求。另外,改进了蚁群算法的信息素更新策略。本文还通过模拟实验测试,证明本文中提出的算法性能高于普通蚁群算法应用于该领域的性能。In order to solve the proposition of services combination aiming at fulfilling the user's requirement and adapting the dynamic and instable character of Web services combination,we put forward an improved ant colony op-timization algorithm.The new algorithm includes two optimizing process.One is local-optimizing process aiming at obtaining local optimum solution,and the other is global-optimizing process aiming at obtaining global optimum solu-tion.So the algorithm can assure the selection of Web services tend to become best as far as possible,meanwhile,satis-fy multi-characteristic of consumer requirement.In addition,the ACO pheromone updating strategy was improved.We give simulate experiment to test the algorithm and prove the improved-ACO algorithm is more effective than ACO in the relevant domain.

关 键 词：服务组合 蚁群算法 用户需求 优化. 

分 类 号：TP393.09[自动化与计算机技术—计算机应用技术]