语义Web服务索引的构建和带有QoS的服务发现

发布时间：2018-12-09 14:27

【摘要】：Web服务经过多年的技术积淀产生了许多成熟的应用模型,期间伴随出现了大量的可用Web服务。随着语义网相关技术的加入,Web服务的描述具有了丰富的语义信息,成为了语义Web服务。本体关系推理工具开始广泛使用于语义Web服务发现过程中,虽然它极大地改善了服务发现算法的精度,但却影响了服务发现的速度。另一方面,由于Web服务数量的增长,服务的安全性问题开始凸显。用户越来越关注服务的可靠性、可信度等QoS因素,并且服务发现算法返回给用户的服务越来越多而且这些服务功能都是相同的,用户需要在这些服务中进一步选择,而Web服务的安全性也就成为他们选择的主要依据。QoS技术是保证Web服务安全性的一个重要手段,因此服务发现模型必须在Web服务安全性、可靠性等QoS参数的描述、匹配与选择方面给用户提供更多地支持,使得衡量广告服务和请求服务安全性、可靠性的QoS属性能够被精确地描述和匹配,力求发现的服务可靠、访问安全、过程可控。针对语义Web服务发现过程中,由于使用推理工具导致匹配速度较慢的问题,本文研究了一种基于索引的服务发现技术并将其应用于服务的功能匹配过程中。首先使用本体概念来描述服务的各项功能参数并利用本体概念间的包含关系来定义服务匹配的必要条件,而本体概念间的包含关系则依据其在本体概念树中的位置来确定。接着对这些本体概念进行区间编码然后建立索引(B+-树或R-树),通过在索引中进行范围查询就可以完成服务功能参数的语义匹配。讨论的最后给出了具体的服务功能匹配算法。对于用户关心的Web服务安全性和可靠性,本文先研究了Web服务QoS的特点并定义了一个QoS本体描述模型—QoSOL,用于对Web服务的安全性、可靠性、性能、网络环境等QoS参数进行精确的描述。在此基础上,本文给出了一个语义Web服务QoS参数的匹配模型。模型要求先从语义角度考察服务提供方和请求方QoS参数度量值是否匹配。只有语义匹配成功,才会使用约束规划方法考察各度量值的数值约束是否匹配,而且模型将对匹配结果进行详细的分类,能够给用户更多的反馈信息。另外,本文还讨论了服务质量评价函数的设计以及基于QoS的Web服务选择。最后通过实验表明,在大规模的语义Web服务发现过程中,使用索引来完成服务参数的语义匹配能够极大的减少搜索空间和比较次数,使得服务发现过程满足一定精度要求的同时更为快速,而使用QoSOL本体和相应的匹配模型对Web服务的安全性、可靠性、性能、网络环境等QoS参数进行语义描述与匹配也是可行的,从而使得Web服务的发现与使用更安全、可靠和灵活。
[Abstract]:After many years of technology accumulation, Web services have produced many mature application models, which are accompanied by a large number of available Web services. With the addition of semantic web related technologies, the description of Web services has abundant semantic information and becomes a semantic Web service. Ontology relational reasoning tool has been widely used in semantic Web service discovery process. Although it greatly improves the accuracy of service discovery algorithm, it affects the speed of service discovery. On the other hand, due to the growth of Web services, the security problems of services begin to highlight. Users pay more and more attention to QoS factors such as reliability and reliability of services, and more services are returned to users by service discovery algorithms, and the functions of these services are the same. QoS technology is an important means to ensure the security of Web services, so the service discovery model must be described in the QoS parameters such as Web service security, reliability and so on. Matching and selection provide users with more support, so that the QoS attributes of reliability and reliability can be accurately described and matched to determine the security of advertising service and request service, and make every effort to find a reliable service, secure access, and controllable process. In order to solve the problem that the matching speed is slow due to the use of inference tools in the process of semantic Web service discovery, this paper studies a service discovery technology based on index and applies it to the process of service function matching. Firstly, the functional parameters of the service are described by ontology concept and the necessary condition of service matching is defined by the inclusion relation between ontology concepts, and the inclusion relation among ontology concepts is determined according to its position in the ontology concept tree. Then, these ontology concepts are intervalently coded and indexed (B-tree or R-tree), and the semantic matching of service function parameters can be accomplished by range query in the index. At the end of the discussion, a specific service function matching algorithm is given. For the security and reliability of Web services concerned by users, this paper first studies the characteristics of Web services QoS and defines a QoS ontology description model-QoSOL, for the security, reliability and performance of Web services. QoS parameters such as network environment are accurately described. On this basis, a matching model of semantic Web service QoS parameters is presented. The model requires that the QoS parameter measures of the provider and the requester be matched from the semantic point of view. Only if the semantic matching is successful, can the constraint programming method be used to investigate whether the numerical constraints of each metric are matched or not, and the model will classify the matching results in detail, which can give users more feedback information. In addition, this paper also discusses the design of QoS evaluation function and Web service selection based on QoS. Finally, experiments show that in the process of large-scale semantic Web service discovery, the use of index to complete semantic matching of service parameters can greatly reduce the search space and the number of comparisons. It makes the service discovery process meet certain precision requirements more quickly at the same time, and it is feasible to use QoSOL ontology and corresponding matching model to describe and match QoS parameters such as security, reliability, performance, network environment and so on. So that the discovery and use of Web services is more secure, reliable and flexible.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2011
【分类号】：TP393.09

【参考文献】