Identifying Vulnerable User In Linkedin Using Web Description Logic Rule Generation

摘要

One of the most preferred networks by professionals amid social networks is LinkedIn. The rapid and explosive growth of these social networks has enabled certain people to misuse the same for illegal and unethical conducts. Nonetheless, considering LinkedIn, these behavioral assertions prove very restrictive in the openly available profile information for users by privacy policies. The publicly present profile information of LinkedIn is limited. Here, it is suggested to pinpoint maximum group of the profile information required to identify vulnerable user in LinkedIn and also determine the proper data mining strategy for this task. In this paper Web Description Logic Rule Generation algorithm is put forth to find and examine vulnerable users and also used to remove the attackers from LinkedIn. Using this algorithm, identifying vulnerable users and to protect them against the attackers is possible according to the sharing threshold. When the threshold value exceeds the limit, the shared person will be removed from OSN. It is demonstrated that using limited profile information, this strategy is capable of spotting attackers at an accuracy of 94% and as low as 3.67% false negative.