Dong Hongbin

Intelligent Viral Marketing Algorithm over Online Social Network

As the online social network become increasingly popular nowadays, performing viral marketing over it has become the focus of many marketing management. How to select a fixed number of initial users from the total population with the purpose of maximizing the profits has long been open as a typical discrete approximation problem. However most of the existing solutions under the setting of online social network tried to traverse every node using network properties which is time-consuming and ineffective. This paper attacks the problem successfully by implementing intelligent algorithms such as GA, DE, PSO. Considering of the huge search space, we sharply decrease the scalability of the network through analyzing the datasets and sampling the data according to a power law property. Experiment results showed that the model we designed for solving viral marketing problem outperform other current search methods.

The self-organizing trust model of the Internetware based on evolutionary game

In order to explore the dynamic process of the trust mechanism of the Internetware entities, this paper proposes a self-organizing trust model of the Internetware by means of the replicated dynamic equation based on the Wright-Fisher process, and analyzes the evolutionary direction and the stability of the Internetware system by establishing the incentive and punitive mechanism. This kind of self-organizing trust model has the function of providing a referenced model for the optimization and stability of the Internetware system, improving the credibility of the Internetware system, and making the system run in a long and stable state.

Bidding strategy in continuous double auction market based on Fuzzy Cognitive Map

In this paper, through analyzing the fuzzy cognitive map inference mechanism and the continuous double auction market bidding process, the bidding process met a causal relationship. We design and analyze a new bidding strategy based on Fuzzy Cognitive Map (Fuzzy Cognitive Map bidding strategy, FCMBS), and build a model for Seller Agents with certain emotions based on FCMBS bidding strategy. FCMBS strategy is divided into three bidding stages, trader agent adopts different strategies depending on different bidding stages. Depending on the input information of market, seller agent model has three different emotional states. In the experiments, we study the performance of the FCMBS bidding strategy and compare it with other classic bidding strategies. Experimental results show that agents adopting the FCMBS bidding strategy were superior to the other strategies in terms of trading success rate and profit.

Mean-Shift moving target tracking algorithm based on weighted sub-block

To reduce the tracking errors caused by the background change and occlusion in dynamic scenes, a novel Mean-Shift moving target tracking algorithm based on weighted sub-block is proposed in this paper. Moving target tracking is completed by blocking the target region. The weight of each sub-block is determined by the combination of the similarity between target sub-block and candidate sub-block and the ratio of the target sub-block area and the overall area. At the same time, the edge position of the target sub-block is found by means of a Sobel operator edge detection algorithm. By which, the target sub-block area is obtained. Both of the target regions RGB color information and the pixels position information are taken into consideration while describing the characteristic model of target and candidate region inside each sub-block. The experiments demonstrate that the proposed method is insensitive to the background change and occlusion, and has better tracking performance with higher tracking accuracy and adaptability.

A Cellular Automaton based Network Diffusion model: Preparation for more scalable Viral Marketing

Nowadays, social network web sites are becoming increasingly popular. Accordingly, viral marketing which aimed to promote products in social network received more and more attention. The Network Diffusion model describes how information or influence is propagated in a social network. It is of great importance to the viral marketing/influence maximization problem. However, there are two challenges: first, how to build the diffusion model so as to simulate the diffusion process better, and second, how improve the efficiency of the diffusion model so that it can be applied to more scalable viral marketing problems. To tackle these two problems, we proposed a Cellular Automaton based Network Diffusion (CAND) model. We also proved that the CAND model is equivalent to the Linear Threshold (LT) model under certain conditions. Moreover, the model is realized and a speed up method is used so as to improve the efficiency of the model. Experiments show that compared with the LT model, our model is much more efficient on four real life social network datasets, indicating a better scalability for Viral Marketing.

A trust evaluation and recommendation algorithm based on Internetware

Internetware is endowed with more flexibility in the dynamic, open network. However, it is facing a big challenge in trust evaluation on interaction between two components. This paper proposes an effective way of trust evaluation based on the punishment mechanism and proved its function to resist malicious attack. As trust relationships can be obtained by the recommendations of other Internetwares, thus a virtual trust information network naturally formed. In this way, the trust relationship of an Internetware can be derived without direct interaction. However, the problem lies in we might get malicious information and be led to misjudgment in such a complex and changeable network. Therefore, we put forward an algorithm based on the idea of RIP routing algorithm. The algorithm is proved to be efficient and reliable through experiment.

Research on Game strategy of online maintenance for Internetware

The paper explores the methods on decision of optimal strategy for online maintenance for Internetware through static Game Theory. An Online Maintenance Game Model (OMGM) is put forward, which simulating the trend of consuming resource and involving the dependent relationship and characteristics of the procedure maintenance among fault Internetwares. In the case of single fault Internetware and double dependent fault Internetwares, we studies calculation methods for comprehensive payoff of various maintenance strategies and extends methods to adopt online maintenance strategy decision for multiple fault Internetwares, thereby, technology and theoretical support are provided to guarantee the scalability and stability of Internetware.