Font Size: a A A

Research On Description Model And Its Compression Mechanism Of Web Services

Posted on:2013-04-21Degree:MasterType:Thesis
Country:ChinaCandidate:Q Y WuFull Text:PDF
GTID:2248330374490008Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
In wired network environment, the technology of Web service discovery is more mature,such as semantic-based service matching techniques and distributed service discoveryarchitecture. However, faced with information and services whether people can quickly obtainor not, that requires further research. In pervasive computing environment, users and providersare expressing mobility and intermittent connectivity, its topology changes every time, andshows a strong dynamic. Moreover, mobile device storage and communication bandwidth arelimited, so the service discovery generated the message load is not too high. In the dynamicnetwork environment, discovering appropriate services is an important prerequisite for shareand reuse, the effects of service discovery are directly related to service reuse quality, andimpact the compatibility and substitutability of service compose, that’s related to whether canquickly use personal service or not. In the service discovery process, the efficient of servicediscovery depends on the service description methods; and pervasive environment relates theservice social relationship, then the paper need to build a social relationship model, in additionto add social property into the service description, and need to compress and optimize thedescription of the service.Firstly, Based on OWL-S ontology language, the paper extends it to a lightweight Webservices description language S-OWL-S. The paper analyzes the social relationship in thepervasive environment, and builds the semantic Web service description model which expressessocial context. Considering the OWL-S service doesn’t describe social context, and it is scalable,so the paper designs the S-OWL-S ontology to describe service language, and build SCProfileontology including the attribute of social context and attribute parameters.Secondly, the paper proposes an improved the Counting Bloom Filter algorithm, which isthe Divide-domain Counting Bloom Filter algorithm. In accordance with the services divisionof the domain, it uses to compress the service information. The purpose of compression issimple to show the set of services, and reduce bandwidth consumption and cache whencommunicate services with users, in addition, it reduces the false positive rate. This paperanalyzes the theory, the algorithm and the false positive rate of the Standard Bloom Filteralgorithm, at the same time, the paper briefly analyzes the Counting Bloom Filter algorithm. Onthe basis of all analysis, the paper proposes the Divide-domain Counting Bloom Filter algorithm,and compares the false positive rate with Bloom Filter, as well as designs hash function. Finally, the paper quantifies the service, and places it to the bit string vector group. BloomFilter algorithm can only express the collection of data, the paper divide service into two partswhich are domain and other service attributes, in accordance with the quantization of thedifferent domain, the paper places service to the different bit string vector group. Usingprecision and recall performance as well as average seek time to check the effectiveness of theDivide-domain Counting Bloom Filter algorithm.
Keywords/Search Tags:pervasive computing, service discovery, social context, service compression, servicequantification, DCBF algorithm
PDF Full Text Request
Related items