上下文感知的Web服务组合优化研究

发布时间：2018-06-13 13:08

本文选题：Web服务组合 + 上下文感知　；参考：《大连海事大学》2014年硕士论文

【摘要】：随着互联网技术的迅猛发展,Web服务技术得到了广泛应用。单个Web服务由于功能单一不能很好地满足用户多样化的需求,因此Web服务组合的概念被提出。当前Web服务组合中存在以下几个问题：首先,由于互联网中存在大量功能相似甚至相同的Web服务,在进行单个Web服务搜索和Web服务组合方案选择时很耗时,导致服务组合效率不高；其次,在进行Web服务组合时,往往只考虑服务的功能需求和部分非功能需求,而没有综合考虑服务的非功能信息、服务之间的关联信息以及用户喜好等,导致服务组合质量不高。此外,在服务组合方案制定的过程中,往往没有考虑服务之间的关联信息,导致服务组合方案在实际运行中成功率不高。针对以上问题,本文将上下文感知技术引入到Web服务组合方案的选择过程中,在Web服务组合过程中充分考虑Web服务及其组合的各种上下文信息,从组合效率、组合质量和组合成功率等方面对Web服务组合进行优化。本文给出了一种上下文感知的Web服务组合优化方法,定义了服务关联上下文的概念,并将服务关联上下文分为统计关联上下文、接口关联上下文和质量关联上下文。在服务组合方案的选择过程中,通过对服务质量上下文、服务关联上下文以及用户上下文等信息的分析,选择出最佳的服务组合方案。首先,将服务统计关联上下文应用到候选服务组合方案集的生成过程中,提高了候选服务组合方案搜索的效率；其次,将服务质量关联上下文和服务接口关联上下文应用到最终服务组合方案的选择过程中,提高了服务组合质量和组合成功率。采用遗传算法对本文研究的上下文感知的Web服务组合优化问题进行了求解,并结合服务关联上下文,对遗传算法的初始化种群、编码方式、遗传操作、适应度函数等方面进行了改进。最后,通过实验数据对比分析,验证了本文给出的上下文感知的Web服务组合优化方法的有效性。
[Abstract]:With the rapid development of Internet technology, Web services technology has been widely used. The concept of Web service composition is proposed because a single Web service can not meet the diverse needs of users. The following problems exist in the current Web service composition: first, because there are a large number of similar or even identical Web services in the Internet, it is time-consuming to search for a single Web service and select a Web service composition scheme. Secondly, in the process of Web service composition, only the functional requirements and some non-functional requirements of the service are considered, and the non-functional information of the service is not synthetically considered. The quality of service composition is not high due to the related information between services and user preferences. In addition, in the process of service composition scheme formulation, the association information between services is often not considered, which leads to the low success rate of service composition scheme in actual operation. Aiming at the above problems, this paper introduces context-aware technology into the process of selecting Web service composition schemes. In the process of Web service composition, we fully consider all kinds of context information of Web services and their composition. Composition quality and composition success rate are optimized for Web service composition. This paper presents a context-aware optimization method for Web service composition, defines the concept of service association context, and classifies service association context into statistical context, interface context and quality context. In the process of selecting the service composition scheme, the best service composition scheme is selected by analyzing the quality of service context, the service association context and the user context. First, the service statistics association context is applied to the generation of candidate service composition scheme set, which improves the efficiency of candidate service composition scheme search. The QoS association context and the service interface association context are applied to the selection process of the final service composition scheme, which improves the service composition quality and the composition success rate. Genetic algorithm is used to solve the context-aware Web service composition optimization problem studied in this paper. Combining with the context of service association, the initialization population, coding mode, genetic operation of genetic algorithm are analyzed. The fitness function is improved. Finally, the effectiveness of the context-aware Web service composition optimization method is verified by comparing and analyzing the experimental data.
【学位授予单位】：大连海事大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.09

【参考文献】