Toward trustworthy social network services: A robust design of recommender systems

Abstract

In recent years, electronic commerce and online social networks (OSNs) have experienced fast growth, and as a result, recommendation systems (RSs) have become extremely common. Accuracy and robustness are important performance indexes that characterize customized information or suggestions provided by RSs. However, nefarious users may be present, and they can distort information within the RSs by creating fake identities (Sybils). Although prior research has attempted to mitigate the negative impact of Sybils, the presence of these fake identities remains an unsolved problem. In this paper, we introduce a new weighted link analysis and influence level for RSs resistant to Sybil attacks. Our approach is validated through simulations of a broad range of attacks, and it is found to outperform other state-of-the-art recommendation methods in terms of both accuracy and robustness.

Social Botnet Community Detection: A Novel Approach based on Behavioral Similarity in Twitter Network using Deep Learning

“…In [12], a trust model is derived by considering Dempster-Shafer theory and Bayes' theorem based on the spam content in Twitter. Social botnet attacks exist in OSNs to manipulate user's (or item's) popularity [15], thereby spreading spam content [29] and creating multiple fake identities [24]. The above existing works rely on the profile-based, network-based, content-based, and click-based features to detect malicious social bots.…”

A robust recommended system based on attack detection

“…6 Since then, the classification of penetration attacks using false contours has been proposed. 2 The work of Gao et al 4 proposed five attack models: sampling attack, random attack, average attack, power flow attack, and staged attack. These five models represent the items sybil wants to work on as a target project.…”

“…Malicious users heavily penetrate RSs and use these counterfeit profiles to attack sybil RS first as in the work of Seminario and Wilson . Since then, the classification of penetration attacks using false contours has been proposed . The work of Gao et al proposed five attack models: sampling attack, random attack, average attack, power flow attack, and staged attack.…”

“…The recommendation system is an intelligent software tool for providing advice to users and has been used in many fields such as e-commerce, movie and video websites, music network radio stations, social networks, as well as personalized reading, mailing and, advertising. [1][2][3] In the increasingly fierce competition in e-commerce sites, the recommendation system has brought great benefits to the site, but some unscrupulous businesses use the recommended system algorithm loopholes, the use of bulk injection of some fake users and the ratings of these users with the normal user's rating. Similarly, the user's proximity circle is easily calculated when calculating the user's similarity level.…”

Trust-Based Video Management Framework for Social Multimedia Networks

“…The providers of these services have enabled countless features that allow users to interact among themselves by creating and sharing different contents (e.g, videos, text, and images). However, by allowing this, the nodes composing the social networks, users and machines, generate a huge amount of data, which can be uncontrolled, unsecured and untrusted [5], [6]. Such amount of generated data are causing a congestion to the networks [7], [8] and posing a new security challenge to the service providers: it becomes hard to handle and analyze all content traversing their networks.…”