مرضيه سروري زاده

The growing popularity and diversity of social media sites have led people to become widespread in online social networks and benefit from their services. Due to the various features of various online social networking systems, users are attracted to these networks for a variety of purposes, such as information search / sharing and storing information. As a user uses various social networks for different purposes, analyzing user identities in a single social media may only give a little understanding of the personality and interests of the individual. However, if we can link the identity of the user to several datasets, we can collect and analyze our data on these social media sites, consequently, we will have a more comprehensive view of the user and we can provide better services. . Each user can create a user ID on each social network, which includes the profile information, the content of the messages and the network information of the friend's friends. The main purpose of this report is to first find the candidate set in the target network for each U_Si user in the source network. After the initial refinement, an unregistered method called the User Linkage based on Fuzzy Logic (ULFL), which performs a candidate based on profile information, , an‎d by comparing similarity in all three profiles, the friends' network and the content of the messages can be found by the fuzzy approach to the best of the users.  The results of this study show that after applying our method, a good improvement in accuracy is obtained. Also, after comparing the results of the method based on all three domains, in comparison with a method based solely on the profile data, we conclude that the use of its small profile data alone provides a very good accuracy.