Multiple Attribute Decision Making Based On The P2p Multi-source Web Service Discovery Research

In recent years, with the rapid development of internet technology, web service as the key technology of service oriented architeture is attracting more and more attention. Web service discovery serves as one of the most important premises to realize web service correct and flexible interaction and web service dynamic composition, it is an urgent problem to be solved that how to efficiently and accurately select service from plenty of web services with the same or similar functions which can satisfy service requestor’s requirement. At the same time, traditional service register center has the centralization bottleneck problem.In order to solve the above problems, implement the web services discovery prototype system based on distributed network after analyzing research on correlation between web services discovery and selection, and on this basis proposes the research on service discovery with certainty multiattribute decision-making based on peer-to-peer and the research on service discovery with uncertainty multiattribute decision-making based on distributed clustering. Main researches as follows:(1)Mainly makes use of Chord distributed peer-to-peer network technology, keeps the services combined together which are grabbled from different service register center, and formed a distributed sharing services center. Develop a service retrieval platform thereon and realize a web services discovery prototype system based on P2P condition, solve the centralization bottleneck and single-point of failue problem of traditional service register center.(2)In order to make the provided web services extremely satisfy for service requestor’s quality of service requirement and improve the service retrieval quality, propose the services discovery research on certainty multiattribute decision-making based on P2P. This research makes the genetic algorithm apply to multiattribute decision-making for picking out the services with high quality of service. Under the circumstance of services quality information determinated, realize the services discovery research of certainty multiattribute decision-making based on P2P. Experiment shown that this method from whole come to the aim at improving service retrieval quality.(3)In order to improve service retrieval efficiency under the P2P condition, propose services discovery research of uncertainty multiattribute decision-making based on distributed clustering. This research realizes clustering processing for distributed services. Meanwhile according to the circumstance in which most services quality are uncertain, as Bayes network is one of the most effective theoretical models, use Bayes network to solve the uncertainty problem which overcomes the subjectivity about exiting measurement method of quality of service. Besides, combine multiattribute decision-making method to select the services which can greatest satisfy service requestor’s requirement, and experiment shown that this method can not only ensure quality of service but also can effectively improve retrieval quality.