With the popularity of location-based services in the field of mobile network applications, users enjoy the convenience on one side, they may face the risk of privacy disclosure on the other side. Attackers can easily dig out the user's home address, occupation, and other personal privacy from the data of location-based services. For the problem of user location privacy disclosure based on semantic query of point of interest (POI) in the road network environment, the previous research seldom paid attention to the temporal association relationship between the query semantic and the user's location, as well, seldom considered whether the constructed anonymity sets of fake-locations can fulfill the property of reciprocity. Therefore, in this paper, we propose a location privacy protection algorithm T-SR based on POI query. The algorithm can be divided into three steps. First, the whole road network is divided into multiple areas based on the Voronoi structure, so that the areas are independent and non-overlapping, and the road grid that contains user's real location will be located according to the level of privacy protection H. Second, the temporal association between the POI semantics are mined from the real check-in data based on Gaussian distribution σ rule, and the POIs which have weak association with the POI of user's location will be filtered to resist the temporal association attack. Third, the POI semantics within the grid will be split into several buckets according to query times record, anonymity degree k, and semantic degree l, then the dummy locations will be selected from various kinds of POIs of the bucket where the user's location is. The resulted anonymity set can defend against replay attack, inference attack, and temporal association attack. The theoretical analysis and experimental evaluation prove that the proposed solution can protect user's location privacy efficiently and effectively.

Recommender system is widely used as an important tool in various fields for effectively dealing with information overload, and collaborative filtering algorithm plays a vital role in the system. However, such system is highly vulnerable to malicious attacks, especially shilling attack because of data openness and independence. Therefore, detecting shilling attack has become an important issue to ensure the security of recommender system. Most of existing methods for detecting shilling attack are based on rating classification features and their limitation is that they are easily to be interfered by obfuscation techniques. Moreover, traditional detection algorithms can not handle multiple types of shilling attack flexibly. In order to solve these problems, in this paper, we propose an outlier degree shilling attack detection algorithm based on dynamic feature selection. By considering the differences of user choosing items and taking user popularity as a detection metric, as well as using information entropy to select detection metrics dynamically, a variety of shilling attack models can be dealt with flexibly. Experiments show that the algorithm has stronger detection performance and interference immunity in shilling attack detection.

No abstract