Showing papers by "Xiangjie Kong published in 2016"

Exploiting Publication Contents and Collaboration Networks for Collaborator Recommendation.

Abstract: Thanks to the proliferation of online social networks, it has become conventional for researchers to communicate and collaborate with each other. Meanwhile, one critical challenge arises, that is, how to find the most relevant and potential collaborators for each researcher? In this work, we propose a novel collaborator recommendation model called CCRec, which combines the information on researchers' publications and collaboration network to generate better recommendation. In order to effectively identify the most potential collaborators for researchers, we adopt a topic clustering model to identify the academic domains, as well as a random walk model to compute researchers' feature vectors. Using DBLP datasets, we conduct benchmarking experiments to examine the performance of CCRec. The experimental results show that CCRec outperforms other state-of-the-art methods in terms of precision, recall and F1 score.

OnCampus: a mobile platform towards a smart campus.

Abstract: An increasing number of researchers and practitioners are working to develop smart cities. Considerable attention has been paid to the college campus as it is an important component of smart cities. Consequently, the question of how to construct a smart campus has become a topical one. Here, we propose a scheme that can facilitate the construction of a smart and friendly campus. We primarily focus on three aspects of smart campuses. These are: the formation of social circles based on interests mining, the provision of educational guidance based on emotion analysis of information posted on a platform, and development of a secondary trading platform aimed at optimizing the allocation of campus resources. Based on these objectives, we designed and implemented a mobile platform called OnCampus as the first step towards the development of a smart campus that has been introduced in some colleges. We found that OnCampus could successfully accomplish the three above mentioned functions of a smart campus.

Cross-domain item recommendation based on user similarity

Abstract: Cross-domain recommender systems adopt multiple methods to build relations from source domain to target domain in order to alleviate problems of cold start and sparsity, and improve the performance of recommendations. The majority of traditional methods tend to associate users and items, which neglected the strong influence of friend relation on the recommendation. In this paper, we propose a cross-domain item recommendation model called CRUS based on user similarity, which firstly introduces the trust relation among friends into cross-domain recommendation. Despite friends usually tend to have similar interests in some domains, they share differences either. Considering this, we define all the similar users with the target user as Similar Friends. By modifying the transfer matrix in the random walk, friends sharing similar interests are highlighted. Extensive experiments on Yelp data set show CRUS outperforms the baseline methods on MAE and RMSE.

Editorial behaviors in peer review

Abstract: Editors play a critical role in the peer review system. How do editorial behaviors affect the performance of peer review? No quantitative model to date allows us to measure the influence of editorial behaviors on different peer review stages such as, manuscript distribution and final decision making. Here, we propose an agent-based model in which the process of peer review is guided mainly by the social interactions among three kinds of agents representing authors, editors and reviewers respectively. We apply this model to analyze a number of editorial behaviors such as decision strategy, number of reviewers and editorial bias on peer review. We find out that peer review outcomes are significantly sensitive to different editorial behaviors. With a small fraction (10 %) of biased editors, the quality of accepted papers declines 11 %, which indicates that effects of editorial biased behavior is worse than that of biased reviewers (7 %). While several peer review models exist, this is the first account for the study of editorial behaviors that is validated on the basis of simulation analysis.

Social-Oriented Resource Management in Cloud-Based Mobile Networks

Abstract: One of the predominant challenges for cloud-based services is the coordination of resources in the application and network layers to provide adaptive service delivery and an acceptable user experience. This article surveys resource management in cloud-based networks. The authors suggest that consideration of users' social associations can improve link connectivity and service delivery, and that user experiences, including both subjective and objective factors, should be integrated in a uniform pattern. They also find that software-defined networking (SDN) technology is promising for resource management in cloud-based networks, allowing different clients to access the network effectively. The authors present some enabling technologies to create a blueprint for user experience-oriented resource management in cloud networks based on SDN technology.

Exploiting Trust and Usage Context for Cross-Domain Recommendation

Abstract: Cross-domain recommender systems are usually able to suggest items, which are not in the same domain, where users provided ratings. For this reason, cross-domain recommendation has attracted more and more attention in recent years. However, most studies propose to make cross-domain recommendation in the scenario, where there are common ratings between different domains. The scenario without common ratings is seldom considered. In this paper, we propose a novel method to solve the cross-domain recommendation problem in such a scenario. We first apply trust relations to the cross-domain scenario for predicting coarse ratings pertaining to cross-domain items. Then, we build a new rating matrix, including known ratings and predicted ratings of items from different domains, and transform a user-item matrix into an item–item association matrix. Finally, we compute the similarities of items belonging to different domains and use item-based collaborative filtering to generate recommendations. Through relevant experiments on a real-world data set, we compare our method to a trust-aware recommendation method and demonstrate its effectiveness in terms of prediction accuracy, recall, and coverage.

New challenges of real-time wireless sensor networks: Theory and applications:

Abstract: Wireless communication is steadily increasing in several applications where the main aim is to reach real-time requirements in terms of quality-of-service (QoS) parameters, bounded latency, high reliability, and dependability, particularly with the advent of HetNet architectures and context-aware environments. The last open research challenges have shown that the application requirements such as energy efficiency, security, and network adaptivity influence the real-time requirements of communication. For this reason, there is a need to disseminate, streamline, and investigate research findings coming from several different application domains. Thus, it is needed to introduce approaches, protocols, architectures, algorithms, and so on, that are able to meet the typical real-time network requirements. The goal of special issue entitled ‘‘New Challenges of Real-Time Wireless Sensor Networks: Theory and Applications’’ is to report on innovative ideas and solutions for the designing of real-time wireless communication in the emerging applications era, focusing on the development, adoption, and application of wireless technology for real-time applications. The special issue received 26 papers, but only 8 have been accepted for publication in International Journal of Distributed Sensor Networks. The paper entitled ‘‘Message Passing Based Time Synchronization in Wireless Sensor Networks: A Survey’’ by Mohammad Ali Sarvghadi and Tat-Chee Wan presents a classification of several Message Passing based Time Synchronization (MPTS) protocols based on various metrics, such as structure formation of the network affected by the synchronization protocol, frequency of synchronization process (synchronization interval), and synchronization message overhead. Moreover, the authors propose some potential methods in order to improve the synchronization process. The paper entitled ‘‘A Flexible and Scalable Architecture for Real-Time ANT+ Sensor Data Acquisition and NoSQL Storage’’ by Nadeem Qaisar Mehmood, Rosario Culmone, and Leonardo Mostarda describes a system architecture based on ANT+, an open access low energy protocol, which has enabled the implementation of a healthcare monitoring system. The proposed solution is scalable and can provide further functionalities in the near future. The paper entitled ‘‘A Performance Analysis of M2M Sensor Networks,’’ by Jingjing Wang, Lingwei Xu, Xinli Dong, Wei Shi, and Qiuna Niu, focuses on the average symbol error probability (ASEP) and outage probability (OP) performance of mobile-to-mobile (M2M) sensor networks employing transmit antenna selection (TAS) and selection combining (SC) over NNakagami fading channels. The exact ASEP and closed-form OP expressions are derived for several modulation schemes, based on the moment generating function (MGF) approach, and the performance under different conditions are evaluated through numerical simulations. The results show that the number of antennas, the fading coefficient, and the number of cascaded components have an important influence on the ASEP and OP performance. The paper entitled ‘‘A Cooperative Beamforming for Physical-Layer Security in Power-Constrained Wireless Sensor Networks with Partial Relay Selection’’ by Mujun Qian, Chen Liu, and Yulong Zou investigates on beamforming schemes in a cooperative wireless sensor network (WSN) for physical-layer security. The authors show that the optimal beamforming scheme should be performed along with a partial relay-selection strategy and then propose two partial relay-selection based beamforming schemes. Simulation results show that the proposed schemes combine the advantage of the all-relay-based scheme in high-power range and that of the best-relay-based scheme in low-power range. The paper entitled ‘‘A Statistical Approach in Designing an RF-Based Human Crowd Density Estimation System’’ by S. Y. Fadhlullah and Widad Ismail proposes a novel technique in order to analyze Human Crowd Density values as a function of ZigBee radio frequency (RF) measurement results. Two different techniques, namely, one-way analysis of variance and design of experiment, have been employed in order to gain insight in the differences between static and dynamic crowds and to identify specific crowd properties as a function of RF parameters. The employed

Crowd Activity Classification Using Category Constrained Correlated Topic Model

Abstract: Automatic analysis and understanding of human activities is a challenging task in computer vision, especially for the surveillance scenarios which typically contains crowds, complex motions and occlusions. To address these issues, a Bag-of-words representation of videos is developed by leveraging in...

Erratum to: OnCampus: a mobile platform towards a smart campus.

Abstract: [This corrects the article DOI: 10.1186/s40064-016-2608-4.].