The Analysis Of Malicious Activity Under The SNS Environment Based On Semantic Web

Social network sites (SNS)has already become a major online application. At the same time, inorder to provide a good user experience for such a large number of users, the social network’s ownfunction also becomes more and more complicated.Interaction from single text information to moredigital content, such as video, etc can all be achieved through this network. So a new challenge forthe social network’s security and privacy protection has been posed.The article first introduced the related problems from the aspects of the status quo of bothdomestic and foreign researches. We expounds the related theory foundation and introduces somerelated principles. Then, we put forward the calculation model of the sharing relationship betweensocial network users based on the ontology and expand the one-to–one relationship between users.By introducing the user relation sharing strategy, we may get the best use of the blacklist filteringmodel which we proposed.To be specific, first of all, the paper proposed an automatically filtering blacklist model of socialnetwork based on semantic web, customized the existing social network ontology model, andproposed a possible construction scheme of the malicious information of the social network.Then,we put forward the solution to the automatically filtering blacklist ontology model by using thesemantic web technology. According to the production of trust in reality, we abstracted the relatedfactors, proposed a reasonable classification standard and by introducing the Pearson algorithm, thetrust between internal and external factors are quantified. And and the same time, a possibleblacklist sharing model based on the user relationship was given on the basis of the blacklistfiltering model we proposed, the latter one is an extension to the former one.At last, simulatedanalysis,which evaluated from both nDCG algorithm and tidalTrust algorithm, was done to provethe validation of the whole theory.