基于用户偏好的Web服务组合方法研究

发布时间：2018-03-14 07:23

本文选题：Web服务　切入点：Web服务组合　出处：《西南大学》2014年硕士论文　论文类型：学位论文

【摘要】：近年来,随着电子商务的迅猛发展,形形色色的Web服务大量的涌现,服务提供商也不断将现存的Web服务整合起来形成新的、增值的服务,去不断的满足用户的需求。不过,用户在各种需求能够得到满足的同时,也对如何才能从这些海量的功能性相同的Web服务集中挑选出符合自身需求的服务感到力不从心。因此,如何有效地挑选出符合用户偏好的Web组合服务,尤其是符合用户特定需求的Web组合服务,渐渐成为学术界研究的热点。而在目前的一些基于QoS的Web服务组合方法研究中,一般采用一些预定的局部约束对各个服务候选集中的原子服务进行筛选,并不注重用户的偏好,对各个子任务的QoS指标以及各个子任务的权重同等处理。这些都是有悖于用户日常生活习惯的。有鉴于此,本文从用户的角度出发,对基于用户偏好的Web服务组合方法做了以下研究工作： (1)基于用户偏好的Web服务组合框架的建立。本文在综合考虑现有的服务组合框架后,提出一个基于用户偏好的服务祝贺框架,增加用户偏好确定模块和服务组合规划制定模块,并注重框架与用户的互动性,更好的去反映用户的偏好。该框架能够将各个服务组合流程模式化,规范整个Web服务的组合流程,更好的满足用户的偏好。 (2)服务组合中全局约束的分解。本文采用FAHP的方法量化用户在各个子任务以及QoS指标上的偏好,设定每个子任务的权重,并根据用户的偏好分解全局约束,将得到的这些局部约束信息和各个子任务权重信息发送到服务选择模块。 (3)满足局部约束的N次筛选流程。服务选择模块中对由大量的功能性相同非功能性不同的Web服务组成的候选服务集,我们采用cskyline算法筛选候选服务集,缩小候选服务集的规模,并采用N-FAHP方法确定用户在各个子任务的QoS指标的权重。在此基础上,我们按照各个子任务权重的大小,采用N次筛选的方法筛选组合服务,尽可能高的去提高用户对Web服务组合规划的满意度。 (4)基于用户偏好服务组合方法的实例分析。由于本文主要侧重于用户偏好的服务选择,与用户的主观性紧密相关,不同的用户由于他们的偏好不同,运行得到的服务结果很可能不同,而且用户对服务组合的满意度也会因人而异,无法通过实验模拟方法的有效性。因此本文借用一个用户出行的实例来验证本方法的可行性和可操作性。
[Abstract]:In recent years, with the rapid development of electronic commerce, a large number of Web services have emerged, and service providers have constantly integrated existing Web services to form new, value-added services to meet the needs of users. Users can not only satisfy their needs, but also find it difficult to pick out services that meet their needs from these massive sets of Web services with the same functionality. How to effectively select the Web composition services that meet the user's preferences, especially the Web composition services that meet the specific needs of the users, has gradually become a hot topic in academic circles. However, in the current research of Web service composition methods based on QoS, In general, some predetermined local constraints are used to filter atomic services in each service candidate set, without paying attention to the preferences of users. The QoS index of each sub-task and the weight of each sub-task are treated equally. These are all contrary to the user's daily life habits. In view of this, this paper starts from the user's point of view, The following research work has been done on the Web service composition method based on user preference:. 1) the establishment of Web service composition framework based on user preference. After considering the existing service composition framework, this paper proposes a service congratulations framework based on user preference, which includes user preference determination module and service composition planning module. The framework can model each service composition process, standardize the whole Web service composition process, and better meet the user's preference. (2) decomposition of global constraints in service composition. This paper uses FAHP method to quantify users' preferences on each subtask and QoS index, sets the weight of each subtask, and decomposes the global constraint according to the user's preference. These local constraint information and each sub-task weight information are sent to the service selection module. In the service selection module, we use the cskyline algorithm to filter the candidate service set, which is composed of a large number of Web services with the same functionality and non-functional, so as to reduce the size of the candidate service set. The N-FAHP method is used to determine the weight of the user's QoS index in each sub-task. On this basis, according to the weight of each sub-task, we use the method of N times to filter the composite service. As high as possible to improve the user's Web services composition planning satisfaction. 4) A case study based on the method of user preference service composition. Because this paper mainly focuses on the service selection of user preference, it is closely related to the subjectivity of users, and different users have different preferences because of their different preferences. The result of running the service is likely to be different, and the satisfaction of the user with the service composition will vary from person to person. The effectiveness of the experimental simulation method can not be achieved, so this paper uses an example of user travel to verify the feasibility and maneuverability of this method.
【学位授予单位】：西南大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.09

【参考文献】