Online social networks have now become the most popular platforms for people to share information with others. Along with this, there is a serious threat to individuals’ privacy. One privacy risk comes from the sharing of co-owned data, i.e., when a user shares a data item that involves multiple users, some users’ privacy may be compromised, since different users generally have different opinions on who can access the data. How to design a collaborative management mechanism to deal with such a privacy issue has recently attracted much attention. In this paper, we propose a trust-based mechanism to realize collaborative privacy management. Basically, a user decides whether or not to post a data item based on the aggregated opinion of all involved users. The trust values between users are used to weight users’ opinions, and the values are updated according to users’ privacy loss. Moreover, the user can make a trade-off between data sharing and privacy preserving by tuning the parameter of the proposed mechanism. We formulate the selecting of the parameter as a multi-armed bandit problem and apply the upper confidence bound policy to solve the problem. Simulation results demonstrate that the trust-based mechanism can encourage the user to be considerate of others’ privacy, and the proposed bandit approach can bring the user a high payoff.

Collaborative filtering is widely used in recommendation systems. A user can get high-quality recommendations only when both the user himself/herself and other users actively participate, i.e., provide sufficient ratings. However, due to the rating cost, rational users tend to provide as few ratings as possible. Therefore, there exists a tradeoff between the rating cost and the recommendation quality. In this paper, we model the interactions among users as a game in satisfaction form and study the corresponding equilibrium, namely satisfaction equilibrium (SE). Considering that accumulated ratings are used for generating recommendations, we design a behavior rule which allows users to achieve an SE via iteratively rating items. We theoretically analyze under what conditions an SE can be learned via the behavior rule. Experimental results on Jester and MovieLens data sets confirm the analysis and demonstrate that, if all users have moderate expectations for recommendation quality and satisfied users are willing to provide more ratings, then all users can get satisfying recommendations without providing many ratings. The SE analysis of the proposed game in this paper is helpful for designing mechanisms to encourage user participation.

With the development of social media technologies, sharing photos in online social networks has now become a popular way for users to maintain social connections with others. However, the rich information contained in a photo makes it easier for a malicious viewer to infer sensitive information about those who appear in the photo. How to deal with the privacy disclosure problem incurred by photo sharing has attracted much attention in recent years. When sharing a photo that involves multiple users, the publisher of the photo should take into all related users' privacy into account. In this paper, we propose a trust-based privacy preserving mechanism for sharing such co-owned photos. The basic idea is to anonymize the original photo so that users who may suffer a high privacy loss from the sharing of the photo cannot be identified from the anonymized photo. The privacy loss to a user depends on how much he trusts the receiver of the photo. And the user's trust in the publisher is affected by the privacy loss. The anonymiation result of a photo is controlled by a threshold specified by the publisher. We propose a greedy method for the publisher to tune the threshold, in the purpose of balancing between the privacy preserved by anonymization and the information shared with others. Simulation results demonstrate that the trust-based photo sharing mechanism is helpful to reduce the privacy loss, and the proposed threshold tuning method can bring a good payoff to the user.

The fast development of vehicle positioning technologies has paved the way for in-vehicle recommendation systems. Successive Point-of-Interest (POI) recommendation, one of the most common forms of in-vehicle recommendation, can help users choose places where they might visit in the future. Generally, to recommend POIs for users, raw check-in data are collected from in-vehicle devices, which raises users’ privacy concerns. To address this problem, we propose a successive POI recommendation framework with local differential privacy. The proposed framework exploits three types of influence factors for recommendation, including geographical influence, temporal influence, and transition pattern between two POIs. To protect users’ check-in data, we propose a distance-weighted location perturbation mechanism that realizes local differential privacy. Considering that different users may have different privacy requirements, we enhance this perturbation mechanism to provide personalized privacy protection. We use the EM reconstruction method to correct the error caused by perturbation on the server side. To improve the accuracy, we design an individual-weighted EM reconstruction method. Simulation results demonstrate that the proposed methods can provide strong privacy protection for users without hurting the recommendation accuracy too much.

With the growing popularity of mobile social networks, point-of-interest (POI) recommendation, which utilizes users’ check-in data to suggest interesting places for users, has attracted much attention in recent years. The check-in data, containing time and location information, are closely related to the user’s personal life. Due to privacy concerns, users are reluctant to share check-in data with the service provider, which causes a negative effect on recommendations. It is important for the user to find a balance between privacy and recommendation quality. In this paper, we consider a POI recommendation scenario where an adversary can access the data that a user reports to the service provider. The user sequentially decides whether to check in for the POI he has visited. A stochastic game model is proposed to analyze the interaction between the user and the adversary. To find a good policy for the user, two value iteration algorithms are applied. The proposed game has a large state set, which makes it difficult for policy learning. To deal with this problem, we use some tricks when implementing the minimax Q-learning algorithm, and a set of neural networks are trained to approximate the Q-functions. To evaluate the performance of the learning algorithms, we conduct a series of simulations by using real-world check-in data. Simulation results show that the proposed learning algorithms can help the user to make good decisions, in the sense that the user can get a high long-term return.

The metaverse is an advanced digital world where users can have interactive and immersive experiences. Users enter the metaverse through digital objects created by extended reality and digital twin technologies. The ownership issue regarding these digital objects can be solved by the blockchain-based non-fungible token (NFT), which is of vital importance for the economics of the metaverse. Users can utilize NFTs to engage in various social and economic activities. However, current NFT protocols expose the owner information to the public, which may contradict with the privacy requirement. In this paper, we propose a protocol NFTPrivate that can realize anonymous and confidential trading of digital objects. The key idea is to utilize cryptographic commitments to hide users’ addresses. By constructing proper zero-knowledge proofs, the owner can initiate privacy-preserving yet publicly verifiable transactions. Illustrative results show that the proposed protocol has higher computation and storage overhead than traditional NFT protocols. While we think this is an acceptable compromise for privacy protection.