Font Size: a A A

Research On Composite Service Selection Method And Key Technology Based On Subgraph Search

Posted on:2021-04-18Degree:DoctorType:Dissertation
Country:ChinaCandidate:J T WuFull Text:PDF
GTID:1368330614461433Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
With the maturity and popularization of Service-Oriented Architecture(SOA),Service-Based System(SBS)has become more and more widely used in service-oriented software engineering.SBS has adopted service composition technology to form a new system that can meet the complex needs of users by finding and combining Web service of multiple loose couplings.At present,service composition technology has become a key and feasible method to build SBS,which has possessed important scientific significance and application value.With the continuously increasing number of Web services,how to use existing service resources to build SBS that can meet user needs has become a current research hotspot.In addition,the invalid single module service of the established SBS operation under dynamic and versatile environment may generate cascade effect,which will cause the whole system can not work normally.Therefore,composite service optimization has appeared to be particularly important to SBS.At present,there have been many researches on service composition technology.However,in the real service selection environment,the traditional method has faced many practical problems.First of all,the existing service composition technology has still required users to possess extent SOA technical knowledge,the SBS built by non-professional users may not be able to be executed normally;Secondly,traditional service composition methods have not considered the diversity of composite service scheme,which has been easy to appear SBS abnormality caused by the failure or timeout of one overlapping component service;In addition,aiming at complex requirements constituted by users,the existing service composition method has been difficult to meet precise requirements of all its component functions and the later fine-tuning and reconstruction have brought huge challenges to users.The thesis has expanded study to the combination service optimization problem of the core in emphasis by centering on problem that how to effectively establish SBS under real environment.Concretely speaking,main work of the thesis has included:(1)The thesis has put forward a kind of composite service optimization method based on previous development experience being aimed to assist systems engineers lacking SOA technical knowledge to build SBS.This method has firstly excavated the cooperation relationship between some services from previous development experience and has generated a Web service data graph;Then,it has converted the service composition problem in the SBS engineering process into a subgraph isomorphism problem;Finally,an improved subgraph isomorphism algorithm is used to provide multiple types of composite service solutions based on Web service data graph to achieve the target SBS.(2)Aiming at the SBS abnormal problem caused by service overlapping during service composition process,the thesis has put forward a kind of diversified Top-k composite service selection method based on greedy mechanism.In this method,the Top-k service composition problem has been modeled as the maximum independent set problem and an efficient pruning strategy and subgraph search algorithm are designed to provide users with diversified composite service solutions.(3)The thesis has firstly tried to conduct modelling to similarity searching problem during SBS engineering process and it has put forward a kind of similarity searching method based on feature.This method has firstly modeled composite service as a group of features and has converted the relaxing process of the task into the feature loss process;Then the thesis has calculated feature loss upper bound based on thought of maximum coverage;Finally,the thesis has filtered the invalid SBS solution according to the feature difference between alternating solution and target SBS,which has provided users with similar composite service solution.This similarity search method does not require time-consuming semantic comparisons,thus opens a new paradigm for efficient SBS engineering that shortens the build cycle.The study of this thesis has solved some practical problems in the selection process of composite services.The work of this thesis has provided an effective and feasible solution to the problem of how to build SBS efficiently and has provided support for the application of carrying out large-scale complex software systems.
Keywords/Search Tags:Service computing, Service composition, Service-based system, Subgraph query, Diversity, Substructure similarity search
PDF Full Text Request
Related items