Heng Luo

A collaborative filtering framework based on both local user similarity and global user similarity

Collaborative filtering as a classical method of information retrieval has been widely used in helping people to deal with information overload. In this paper, we introduce the concept of local user similarity and global user similarity, based on surprisal-based vector similarity and the application of the concept of maximin distance in graph theory. Surprisal-based vector similarity expresses the relationship between any two users based on the quantities of information (called surprisal) contained in their ratings. Global user similarity defines two users being similar if they can be connected through their locally similar neighbors. Based on both of Local User Similarity and Global User Similarity, we develop a collaborative filtering framework called LS&GS. An empirical study using the MovieLens dataset shows that our proposed framework outperforms other state-of-the-art collaborative filtering algorithms.

Cooperativeness prediction in P2P networks

Peer-to-peer systems are open communities, in which not only is there no overarching control, but neither is there any hierarchy of control among the system components. In such open communities where peers can join and leave freely and behave autonomously, selecting appropriate peers to cooperate with is a challenging problem, since the candidate peers may be unreliable or dishonest. Reputation systems have been proposed to boost trust and enhance collaboration among peers. However, conventional computational reputation systems tend to generate trust based on ad hoc aggregation techniques thus produce reputation values with ambiguous meanings. In this paper we propose a probabilistic computational approach to model and generate reputation. By explicitly separating the reputation between providing services and giving recommendations, our solution represents the estimate of service quality for a specific transaction as a probability conditioned upon each retrieved recommendation, thus taking the innate behaviours of reporters into account. A Kalman filter is applied to learn further the service reputation from the estimate. The proposed approach works well even when there is sparse feedback from the reporting peers giving output with well-defined semantics and useful meanings.

Provisioned placement of directional sensors for covering stationary targets

Directional sensor networks have been increasingly deployed in various contexts or environments such as banks, warehouses and museums. Directional sensors, in contrast to normal omnidirectional sensors, are characterized with limited sensing angles, and their sensing region is thus usually modeled by sectors. This paper considers the provisioned deployment of directional sensors for covering a fixed number of stationary discrete targets in a two-dimensional field. Minimizing the number of directional sensors used is desirable for financial concerns. Such minimization is formulated as an Integer Linear Programming problem, which is then proved to be NP-complete. Subsequently, three approximation algorithms named Induction, LP-based and Randomized are proposed in turn. Finally simulation results are given to compare the performance of the three approximation algorithms as well as the classic greedy algorithm.

Metadata-based delivery framework for P2PTV

P2PTV refers to television media streams being distributed to large numbers of viewers on the Internet via peer-to-peer overlay networks. In the design of P2PTV systems, there is a fundamental trade-off between the better Quality-of-Service (QoS) desired by the viewers and the lower resource consumption preferred by the P2PTV distributors as well as the Internet Service Providers (ISPs). This trade-off is partially caused by the time-varying and limited capabilities of the overlay connections. Observing the potential duplication of the television media streams, we propose a metadata-based delivery framework for P2PTV. In contrast to previous efforts, this framework distributes the original media streams as well as the associated metadata information simultaneously. As a result, the streaming clients at the viewer side are able to identify duplicated content segments in media streams and avoid requesting the duplicated content multiple times. In this way, the P2PTV distributors, ISPs and viewers could reduce their upload bandwidth consumption. Moreover, the distributors within this framework can discriminate various media content according to time constraints such that they could apply different uploading policies to live, pre-recorded and hybrid media content. The streaming clients are even allowed to slightly time-shift local playback schedules such that viewers have an opportunity to experience better P2PTV services.