Link Privacy Research Articles

An Uncertain Graph Method Based on Node Random Response to Preserve Link Privacy of Social Networks

An Uncertain Graph Method Based on Node Random Response to Preserve Link Privacy of Social Networks

Preserving Link Privacy in Uncertain Directed Social Graphs With Formal Guarantees

Preserving Link Privacy in Uncertain Directed Social Graphs With Formal Guarantees

A privacy preserving recommendation and fraud detection method based on graph convolution

<abstract> <p>As a typical deep learning technique, Graph Convolutional Networks (GCN) has been successfully applied to the recommendation systems. Aiming at the leakage risk of user privacy and the problem of fraudulent data in the recommendation systems, a Privacy Preserving Recommendation and Fraud Detection method based on Graph Convolution (PPRFD-GC) is proposed in the paper. The PPRFD-GC method adopts encoder/decoder framework to generate the synthesized graph of rating information which satisfies edge differential privacy, next applies graph-based matrix completion technique for rating prediction according to the synthesized graph. After calculating user's Mean Square Error (MSE) of rating prediction and generating dense representation of the user, then a fraud detection classifier based on AdaBoost is presented to identify possible fraudsters. Finally, the loss functions of both rating prediction module and fraud detection module are linearly combined as the overall loss function. The experimental analysis on two real datasets shows that the proposed method has good recommendation accuracy and anti-fraud attack characteristics on the basis of preserving users' link privacy.</p> </abstract>

Large-Scale Social Network Privacy Protection Method for Protecting K-Core

Social network analysis has many important applications and methods which depend on the sharing and publishing of graphs. For example, link privacy requires limiting the probability of an adversary identifying a target sensitive link between two individuals in the published social network graph. However, the existing link privacy protection methods have low processing power for large-scale graph data and less consideration of community protection in the publishing graphs. Therefore, aiming at sensitive link privacy protection, a large-scale social network privacy protection model to protect K-Core (PPMPK) was proposed. The large-scale social network graph was processed to ensure that the core number and the community structure of the nodes were unchanged based on the Pregel parallel graph processing model. Extensive experiments on the real data sets showed that the proposed method could effectively process the large-scale graph data and protect the data availability of the published graphs, especially in community protection.

The new Hungarian asset management foundation

Abstract The most important global issues of wealth planning are privacy, asset protection, and tax compliance. The ultimate question is how these needs match with the current and future legal environment. Obviously, privacy is under attack and the implementation of the fifth Anti-Money Laundering Directive is sweeping away the remaining ruins of it. The separation of private and business assets is one of the bases of the legal work and, during the last 10 years, tax compliance was forced out by The Organisation for Economic Co-operation and Development (OECD) Member States, achieving the result that assets are declared and taxed. There is very little room for manoeuvre, if someone wants to achieve all three aims at once. The new Hungarian asset management foundation—a hybrid of a trust and a private foundation—provides an excellent solution to link privacy, asset protection, and tax compliance together.

Social Network Sensitive Area Perturbance Method Based on Firefly Algorithm

Social network analysis has many essential applications, but it depends on sharing and publishing the underlying graph. The real dataset published by social networks has great appeal to researchers and research institutions. However, publishing datasets can create many security and privacy problems. A standard technique for achieving link privacy is to randomize a link over the space for node pairs based on probabilistically. Aiming at the problem of social network link privacy, a Sensitive Area Perturbance Based on Firefly (SAPBF) algorithm is proposed. Improve the data availability of the published graph by limiting the range of random perturbations. The specific method is to measure the influence of nodes, such as k-core, degree, and PageRank algorithms to obtain nodes with different influences. The firefly algorithm is applied to the social network to find high-influence nodes. The low-influence nodes are gathered around the high-influence nodes to form sensitive areas. Finally, according to different edge retention probability, the random perturbation algorithm protects privacy in the sensitive area. The content of the social network is continuously enriched and the scale is increasing. The performance of the anonymous algorithm of the single-machine social network graph is limited. The SAPBF algorithm is parallelized based on the Pregel model. The results of the algorithm in the real experimental dataset show that the proposed algorithm guarantees the property of the graph while ensuring the anonymity. Compared with the traditional random perturbation algorithm, the distributed processing graph improves the efficiency while ensuring the data availability of the graph.

Differentially private graph-link analysis based social recommendation

Modern social networks always require a social recommendation system which recommends nodes to a target node based on the existing links originate from this target. This leads to a privacy problem since the target node can infer the links between other nodes by observing the recommendations it received. As a rigorous notion of privacy, differential privacy has been used to define the link privacy in social recommendation. However, existing work shows that the accuracy of applying differential privacy to the recommendation is poor, even under an unreasonable privacy guarantee. In this paper, we find that this negative conclusion is problematic due to an overly-restrictive definition on the sensitivity. We propose a mechanism to achieve differentially private graph-link analysis based social recommendation. We make experiments to evaluate the privacy and accuracy of our proposed mechanism, the results show that our proposed mechanism achieves a better trade-off between privacy and accuracy in comparison with existing work.

Neighborhood randomization for link privacy in social network analysis

Social network analysis has many important applications but it depends on sharing and publishing the underlying graph. Link privacy requires limiting the ability of an adversary to infer the presence of a sensitive link between two individuals in the published social network graph. A standard technique for achieving link privacy is to probabilistically randomize a link over the space for node pairs. A major drawback of such graph-wise randomization is that it ignores the structural proximity of nodes, thus, alters considerably the structure of social networks and distorts the accuracy of social network analysis. To address this problem, we propose a structure-aware randomization scheme, called neighborhood randomization. This scheme models a social network as a directed graph and probabilistically randomizes the destination of a link within a local neighborhood. By confining the randomization to a local neighborhood, this scheme drastically reduces the distortion to the graph structure yet hides a sensitive link. The trade-off between privacy and utility is dictated by the retention probability of a destination and by the size of the randomization neighborhood. We conduct extensive experiments to evaluate this trade-off using real life social network data.

On link privacy in randomizing social networks

Many applications of social networks require relationship anonymity due to the sensitive, stigmatizing, or confidential nature of relationship. Recent work showed that the simple technique of anonymizing graphs by replacing the identifying information of the nodes with random IDs does not guarantee privacy since the identification of the nodes can be seriously jeopardized by applying subgraph queries. In this paper, we investigate how well an edge-based graph randomization approach can protect sensitive links. We show via theoretical studies and empirical evaluations that various similarity measures can be exploited by attackers to significantly improve their confidence and accuracy of predicted sensitive links between nodes with high similarity values. We also compare our similarity measure-based prediction methods with the low-rank approximation-based prediction in this paper.