SybilBelief: A Semi-Supervised Learning Approach for Structure-Based Sybil Detection

Gong, Neil Zhenqiang; Frank, Mario; Mittal, Prateek

doi:10.1109/tifs.2014.2316975

Cited by 162 publications

(107 citation statements)

References 35 publications

Supporting

Mentioning

105

Contrasting

Unclassified

Order By: Relevance

“…In addition to that, some other works have adopted a hybrid approach combining both topological (i.e., graph‐based) properties and feature‐based classification. Examples are the works of M. Jiang et al () and Laleh, Carminati, and Ferrari (), or the work of Gong, Frank, and Mittal (), where a semi‐supervised classification methodology starting from sets of known honest and sybil nodes was exploited. Another example is the work of Li, Martinez, Chen, Li, and Hopcroft (), where the authors deploy a semi‐supervised process that starts from knowledge on spammer seeds, and tracks the activities and interactions of users with shared videos in Youtube over time by analyzing temporal graphs.…”

Section: Knowledge‐based Defense Mechanismsmentioning

confidence: 99%

Knowledge‐based approaches for identity management in online social networks

Bahri¹,

Carminati²,

Ferrari³

2018

WIREs Data Min & Knowl

View full text Add to dashboard Cite

When we meet a new person, we start by introducing ourselves. We share our names, and other information about our jobs, cities, family status, and so on. This is how socializing and social interactions can start: we first need to identify each other. Identification is a cornerstone in establishing social contacts. We identify ourselves and others by a set of civil (e.g., name, nationality, ID number, gender) and social (e.g., music taste, hobbies, religion) characteristics. This seamlessly carried out identification process in face‐to‐face interactions is challenged in the virtual realms of socializing, such as in online social network (OSN) platforms. New identities (i.e., online profiles) could be created without being subject to any level of verification, making it easy to create fake information and forge fake identities. This has led to a massive proliferation of accounts that represent fake identities (i.e., not mapping to physically existing entities), and that poison the online socializing environment with fake information and malicious behavior (e.g., child abuse, information stealing). Within this milieu, users in OSNs are left unarmed against the challenging task of identifying the real person behind the screen. OSN providers and research bodies have dedicated considerable effort to the study of the behavior and features of fake OSN identities, trying to find ways to detect them. Some other research initiatives have explored possible techniques to enable identity validation in OSNs. Both kinds of approach rely on extracting knowledge from the OSN, and exploiting it to achieve identification management in their realms. We provide a review of the most prominent works in the literature. We define the problem, provide a taxonomy of related attacks, and discuss the available solutions and approaches for knowledge‐based identity management in OSNs. This article is categorized under: Fundamental Concepts of Data and Knowledge > Human Centricity and User Interaction Application Areas> Internet and Web‐Based Applications Application Areas> Society and Culture

show abstract

Section: Knowledge‐based Defense Mechanismsmentioning

confidence: 99%

Knowledge‐based approaches for identity management in online social networks

Bahri¹,

Carminati²,

Ferrari³

2018

WIREs Data Min & Knowl

View full text Add to dashboard Cite

show abstract

“…[12] proposes a fuzzy C-means method to softly cluster the probability of a user being malicious or benign by examining the deterministic sensing reports in a period. In [13], Sybil detection is determined by exploring the existence of users and their connection information. A semi-supervised learning approach is considered with a joint distribution probability over all pre-defined labels of benign or malicious users.…”

Section: Related Workmentioning

confidence: 99%

Channel-Based Sybil Detection in Industrial Wireless Sensor Networks: A Multi-Kernel Approach

Cheffena

Shen

2017

GLOBECOM 2017 - 2017 IEEE Global Communications Conference

View full text Add to dashboard Cite

Abstract-Industrial Wireless Sensor Networks (IWSNs) integrate various types of sensors to measure and control industrial production. However, the unattended open environment makes IWSNs vulnerable to malicious attacks, such as Sybil attacks, which may degrade the network performance. In addition, multipath distortion, impulse noise and interference effects in the harsh industrial environment may influence the accuracy of attack detection. In this paper, we propose a Sybil detection scheme based on power gain and delay spread analysis by exploiting the spatial variability from their channel responses. Specifically, we utilize channel-vectors to represent the sensor features based on the power gain and delay spread extracted from channel response. Furthermore, we develop a kernel-oriented method to distinguish Sybil attackers from benign sensors by clustering the channel-vectors. In addition, to alleviate the impact of industrial noise and interference effects, we design a multi-kernel based fuzzy c-means method to map the extracted channel-vectors into a new feature space such that the dispersive effects on the channelvectors can be reduced. We also propose a parameter selection method to optimize the employed kernels. The simulation results show that the proposed multi-kernel scheme can achieve high accuracy in detecting the packets from Sybil attackers, and tolerate the dispersive attenuation and interference effects in the industrial environments.

show abstract

“…Then, SybilBelief propagates the label information from the known benign and/or Sybil nodes to the remaining nodes in the system. 23 Compared with above methods, GroupFound is a lightweight algorithm. It does not need training data in advance.…”

Section: Social Feature-based Machine Learning Classifiersmentioning

confidence: 99%

GroupFound: An effective approach to detect suspicious accounts in online social networks

Feng

Pan

et al. 2017

International Journal of Distributed Sensor Networks

View full text Add to dashboard Cite

Online social networks are an important part of people's life and also become the platform where spammers use suspicious accounts to spread malicious URLs. In order to detect suspicious accounts in online social networks, researchers make a lot of efforts. Most existing works mainly utilize machine learning based on features. However, once the spammers disguise the key features, the detection method will soon fail. Besides, such methods are unable to cope with the variable and unknown features. The works based on graph mainly use the location and social relationship of spammers, and they need to build a huge social graph, which leads to much computing cost. Thus, it is necessary to propose a lightweight algorithm which is hard to be evaded. In this article, we propose a lightweight algorithm GroupFound, which focuses on the structure of the local graph. As the bi-followers come from different social communities, we divide all accounts into different groups and compute the average number of accounts for these groups. We evaluate GroupFound on Sina Weibo dataset and find an appropriate threshold to identify suspicious accounts. Experimental results have demonstrated that our algorithm can accomplish a high detection rate of 86:27% at a low false positive rate of 8:54%.

show abstract

SybilBelief: A Semi-Supervised Learning Approach for Structure-Based Sybil Detection

Abstract: Abstract-Sybil attacks are a fundamental threat to the security of distributed systems. Recently, there has been a growing interest in leveraging social networks to mitigate Sybil attacks.

Cited by 162 publications

References 35 publications

Knowledge‐based approaches for identity management in online social networks

Knowledge‐based approaches for identity management in online social networks

Channel-Based Sybil Detection in Industrial Wireless Sensor Networks: A Multi-Kernel Approach

GroupFound: An effective approach to detect suspicious accounts in online social networks

Contact Info

Product

Resources

About