Controllable Information Sharing for User Accounts Linkage across Multiple Online Social Networks

Shen, Yilin; Jin, Hongxia

doi:10.1145/2661829.2661985

Cited by 26 publications

(9 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Semi-supervised approach considers both labeled and unlabeled pairs of matching user identities in model learning [1,2,8,13,18,20,23,24]. HYDRA is a semi-supervised framework which models user behaviors and structure consistency [8].…”

Section: Supervised and Semi-supervised Approachesmentioning

confidence: 99%

Retrofitting Embeddings for Unsupervised User Identity Linkage

Zhou

Lim

Lee

et al. 2020

Advances in Knowledge Discovery and Data Mining

View full text Add to dashboard Cite

User Identity Linkage (UIL) is the problem of matching user identities across multiple online social networks (OSNs) which belong to the same person. The solutions to UIL problem facilitate cross-platform research on OSN users and enable many useful applications such as user profiling and recommendation. As the UIL labeled data are often lacking and costly to obtain, learning user embeddings for matching user identities using an unsupervised approach is therefore highly desired. In this paper, we propose a novel unsupervised UIL framework for enhancing existing user embedding-based UIL methods. Our proposed framework incorporates two key ideas, user-discriminative features and retrofitting embedding. The user-discriminative features enable us to differentiate a specific user identity from other users in its OSN. From the userdiscriminative features, we derive pairs of similar user identities across OSNs for retrofitting the base user embeddings of existing UIL methods. Through extensive experiments on three real-world OSN datasets, we show that our framework can leverage user-discriminative features to improve the accuracy of different user embedding-based UIL methods significantly. The quantum of improvement can also be surprisingly good even for existing UIL methods with very poor matching accuracy. Electronic supplementary materialThe online version of this chapter (https://

show abstract

Section: Supervised and Semi-supervised Approachesmentioning

confidence: 99%

Retrofitting Embeddings for Unsupervised User Identity Linkage

Zhou

Lim

Lee

et al. 2020

Advances in Knowledge Discovery and Data Mining

View full text Add to dashboard Cite

show abstract

“…By mathematically proving the effectiveness of these similarities, a two-stage iterative algorithm CPCC is used for the alignment. Shen and Jin [27] define a distance between the features of users' profile and neighbors', then a propagation algorithm is adopted for iteratively identifying users across networks. Zhong et al [28] propose a weekly supervised alignment model CoLink which incorporates an attribute based model and a relation based model into a co-training style.…”

Section: Related Work a User Alignment Across Social Networkmentioning

confidence: 99%

ABNE: An Attention-Based Network Embedding for User Alignment Across Social Networks

Liu

Zhang²,

et al. 2019

IEEE Access

View full text Add to dashboard Cite

User alignment across social networks can facilitate more information/knowledge transferring across networks and thereby benefit several applications, including social link prediction, cross-domain recommendation, and information diffusion. Several works try to learn a common subspace for networks by preserving the structural proximities, such that different contribution weights of neighbors are ignored as users were always connected by unweighted edges. In this paper, we propose an attention-based network embedding model that exploits the social structures for user alignment. In particular, two main components are contained in our model framework: a masked graph attention mechanism which tries to learn the alignment task driven attention weights by the supervision of pre-aligned user pairs, and an embedding algorithm tries to learn a common vector space by explicitly modeling the weighted contribution probabilities between follower-ships and followee-ships. With the learned weights and embeddings transferring between these two components, we construct a unified model for user embedding and alignment. Stochastic gradient descent and negative sampling are adopted for efficient learning and scalability. The extensive experiments on real-world social network data sets demonstrate the effectiveness and robustness of the proposed model compared with several state-of-the-art methods. INDEX TERMS Attention mechanism, network embedding, social network analysis, user alignment.

show abstract

“…This means a user may hold multiple profiles, which are stored and shared in different SNS [23-25]. For example, a user creates an account on Facebook mainly to communicate with friends and families, as well as to share pictures and videos with them.…”

Section: Privacy Threatsmentioning

confidence: 99%

“…Third parties and companies may use different user accounts and their social relations to connect multiple social networks and produce aggregated user profiles [24,25]. These aggregated user profiles would be immensely valuable to companies looking to market products or services or, in the case of employers, screen potential job applicants.…”

Section: Privacy Threatsmentioning

confidence: 99%

“…For example, malicious third parties and identity thieves may use their own crawler systems to obtain a user’s private information and friend lists. More seriously, such third parties and individuals could create fake accounts pretending to be this person and then solicit others to connect [25]. These fake accounts can be abused to deliberately leak the user’s private information and friend lists to malicious intruders, which could quickly turn into identity theft and fraud, losing a job, hurting relationships, or even worse.…”

Section: Privacy Threatsmentioning

confidence: 99%

See 1 more Smart Citation

A Privacy Preservation Model for Health-Related Social Networking Sites

2015

J Med Internet Res

View full text Add to dashboard Cite

The increasing use of social networking sites (SNS) in health care has resulted in a growing number of individuals posting personal health information online. These sites may disclose users' health information to many different individuals and organizations and mine it for a variety of commercial and research purposes, yet the revelation of personal health information to unauthorized individuals or entities brings a concomitant concern of greater risk for loss of privacy among users. Many users join multiple social networks for different purposes and enter personal and other specific information covering social, professional, and health domains into other websites. Integration of multiple online and real social networks makes the users vulnerable to unintentional and intentional security threats and misuse. This paper analyzes the privacy and security characteristics of leading health-related SNS. It presents a threat model and identifies the most important threats to users and SNS providers. Building on threat analysis and modeling, this paper presents a privacy preservation model that incorporates individual self-protection and privacy-by-design approaches and uses the model to develop principles and countermeasures to protect user privacy. This study paves the way for analysis and design of privacy-preserving mechanisms on health-related SNS.

show abstract

Controllable Information Sharing for User Accounts Linkage across Multiple Online Social Networks

Cited by 26 publications

References 26 publications

Retrofitting Embeddings for Unsupervised User Identity Linkage

Retrofitting Embeddings for Unsupervised User Identity Linkage

ABNE: An Attention-Based Network Embedding for User Alignment Across Social Networks

A Privacy Preservation Model for Health-Related Social Networking Sites

Contact Info

Product

Resources

About