Semantic Web-based Machine Learning Algorithms And Applications

The World Wide Web is being developed into Semantic Web.The goal of Semantic Web is to make the computer understand and process the data which can only be shown by the current Web,to provide various intelligent services etc automantically.The purpose of research in this paper is to make the computer process the information on the Semanic Web intelligently.For this purpose, machine learning algorithms based on semantic web have been done further researches into in this paper.Firstly,an approach for optimizing the rules generated by the ID3 algorithm is presented after the ID3 algorithm of decision tree learning algorithms is discussed in detail.The optimized ID3 algorithm is then applied into an XML document about books sale.The result of running the algorithm shows that attribute values in the XML document are classified correctly,and two generated rules are incorporated into one.The optimized ID3 algorithm can process the XML document intelligently,generate the decision rules,and optimize the rules in the limit range.Secondly,this paper compared the CBR and RBR,and a new integrated reasoning approach is presented.Then a tourism service system,in which case description is written by ontology description language - OWL, is designed and realized based on this approach. The result of system running shows that the system first search similar cases to the custom's case in the case library.It is the process of CBR.If no matching case exists,the process of RBR begins .Finally,the result of RBR is supplied with the customer.This approach integrates advantages of CBR and RBR,and can accomplish intelligent ontology reasoning...