RDF Data Trust Evaluation Based On Evaluation Model And Page Rank Algorithm

With the development of the Web,large numbers of RDF data were published on it.It has become more and more necessary for data consumers to be able to access the trustworthiness of the data,so the trustworthy issue of RDF data has become a hot spot in Web research area.In this thesis,we conduct the trust evaluation of RDF data by integrating RDF content,provenance information and link structure between RDF data.The main researchers in this thesis can be summarized as following:1)By analyzing existing data authority evaluation model RTEM which based on Naming Authority and DING algorithm,problems in the model are raised;2)Based on RTEM,the RDF trust evalution modelTRTEM is built according to the RDF data content;3)Based on TRTEM,the RDF data trust evaluation model PTRTEM is constructed by using the provenance information of RDF data;4)Based on PTRTEM,the trust evaluation model QPTRTEM based on the factors driven method and the QPR algorithm is built;5)The validity of the QPTRTEM model is verified by experiments and the different models are compared and analyzedThe main contributions of this thesis are:1)Introducing the trust mechanisms and combined with the concept of semantic network to construct the trust relations between entities,the method to evaluate RDF data trust based on semantic social network is proposed;2)Considering the trust of the RDF data content itself,the trust calculation which made trust evaluation more reliable by combining provenance information is put forward;3)Combing with the existing Hartig Provenance model,this thesis propose a factor evaluation strategy and improve the existing PageRank data ranking algorithm by putting forward the QPR algorithm which used for filtering the untrusted data sets.