Matching Algorithm Research Of Semantic Web Services Based On Ontology

Its main purposes of semantic web is extending current web,adding semantic markups to information on the web with ontology technology , that can be interpreted and processed by computer automatically. Semantic web will help computer and human collaboration more easily and being as a universal medium for information sharing and exchanging.Dynamic behavior on the semantic web (semantic web services) and its related application technology are one of research hotspots of in recent years. Services matching algorithm is a primary issue in semantic web services that is how to find the exact service which match with user's requirement. With the rapid development of internet, more and more web services are available on the web. Now, the web has become more widespread with exponential growth of online information, users are face with the challenge to search, access, and maintain information which becomes more difficultly. How to find the exact web services quickly and accurately for users is an increasingly critical issue today.To solve this problem, firstly, needs to build a good ontology database and select a good web services matching algorithm for semantic web service. Semantic web service discovery is: using necessary discovery mechanism to achieve the web automatically discovery right web services. Service matching algorithm is the key issue within service discovery that is decide service requirement of requesters and published service description of service providers matching or not, when a published service description and a request's service description matching exactly, that is a perfect match, however, services with those strict definition of search recall rate and precision rate are very low, even lead to failure. Then, it is necessary to find a good service matching algorithm to calculate similarity in order to meet the requirements of users.This thesis makes use of data warehouse and data mining techniques, building a domain ontology as a reference and guiding principles by the central idea of law, the final step is building a good ontology model, setting up a semantic web service matching algorithm with ontology and tree pruning principle to get a new kind of semantic web service matching algorithm. The algorithm solves the problem that how computer can quickly find out right services according to users requests within the large number of Web services on the semantic web. This thesis puts forward a concrete method to build ontology which reduces the amount of useless information in a certain extent. The services matching algorithm of this thesis has advantages such as: ontology constructing is simply and clearly, has good classified capability, has good predictive ability and easily documented. Matching semantic web services with ontology-based decision tree enables services precision rate and recall rate improved that could find compatible services according to user requirements.