Social Network Analysis

Transactions of the Institute of British Geographers

Matrix Factorization Techniques for Recommender Systems

本文对国际科学计量学杂志《Ｓｃｉｅｎｔｏｍｅｔｒｉｃｓ》１９７９－１９９１年的研究论文内容、栏目、作者及国别和编委及国别作了计量分析，揭示出科学计量学研究的重点、活动的中心及发展趋势，说明了学科带头人在发展科学计量学这门新兴学科中的作用。

国际刊物《Scientometrics》文献计量研究

Migrating computational intensive tasks from mobile devices to more resourceful cloud servers is a promising technique to increase the computational capacity of mobile devices while saving their battery energy. In this paper, we consider an MIMO multicell system where multiple mobile users (MUs) ask for computation offloading to a common cloud server. We formulate the offloading problem as the joint optimization of the radio resources—the transmit precoding matrices of the MUs—and the computational resources—the CPU cycles/second assigned by the cloud to each MU—in order to minimize the overall users’ energy consumption, while meeting latency constraints. The resulting optimization problem is nonconvex (in the objective function and constraints). Nevertheless, in the single-user case, we are able to compute the global optimal solution in closed form. In the more challenging multiuser scenario, we propose an iterative algorithm, based on a novel successive convex approximation technique, converging to a local optimal solution of the original nonconvex problem. We then show that the proposed algorithmic framework naturally leads to a distributed and parallel implementation across the radio access points, requiring only a limited coordination/signaling with the cloud. Numerical results show that the proposed schemes outperform disjoint optimization algorithms.

Joint Optimization of Radio and Computational Resources for Multicell Mobile-Edge Computing

As an important branch of the Internet of Things (IoT), the Internet of Vehicles (IoV) has attracted extensive attention in the research field. To deeply study the IoV and build a vehicle spatiotemporal interaction network, it is necessary to use the trajectory data of private cars. However, due to privacy and security protection policies and other reasons, the data set of private cars cannot be obtained, which hinders the research on the social attributes of vehicles in the IoV. Most of the previous work generated the same type of data, and how to generate private car data sets from various existing data sets is a huge challenge. In this paper, we propose a tri-layer framework to solve this problem. First, we propose a novel region division scheme that considers detailed inter-region relations connected by traffic flux. Second, a new spatial-temporal interaction model is developed to estimate the traffic flow between two regions. Third, we devise an evaluation pipeline to validate generation results from microscopic and macroscopic perspectives. Qualitative and quantitative results demonstrate that the data generated in heavy density scenarios can provide strong data support for downstream IoV and mobility research tasks.

RMGen: A Tri-Layer Vehicular Trajectory Data Generation Model Exploring Urban Region Division and Mobility Pattern

Public transport is of great significance in megacities. Transit-oriented development (TOD) has become a reliable solution to urban sustainable development, which can reshape the urban form and improve its quality. This paper focuses on leveraging heterogeneous mega urban data to answer three critical questions in TOD: what region looks like under TOD concept, which regions have the potential to be TOD regions, and how to construct these TOD regions. For region partition, we propose a connected component-based clustering algorithm, which merges the large amount of public transport stops into representative cluster ones as region centers, and then apply the Voronoi algorithm to locate the region boundaries according to the cluster centers. For TOD region identification, we present a link importance-based random walk method that considers the importance of various transits and further identifies the most valuable regions to be TOD. For discovering functions of TOD regions, we introduce a multifactor-based function characterization approach that combines both the static linguistic factor and human mobility factor together and then derives the actual function distributions. The experiments, which are conducted on three real data sets, show the superiority of the proposed methods to solve the problems of region partition, TOD region identification, and function characterization for the megacities. In the meantime, the results provide support for the government to formulate public policy to construct a TOD city.

Discovering Transit-Oriented Development Regions of Megacities Using Heterogeneous Urban Data

Taxis are one of the representative modes of traffic systems. However, with the emergence of shared cars led by DiDi and Uber in recent years, the traditional taxi companies are facing unprecedented competitions. Without personalized data collected from the mobile devices, passenger flow prediction based on vehicle GPS records presents a unique solution that can improve taxis’ operating efficiency while preserving personal privacy. In this paper, we propose the Travel Behavioral Inertia (TBI) from taxi GPS records, which embodies Driver Inertia (DI) and Passenger Inertia (PI). Then we integrate TBI with other features to construct multi-dimensional features and predict taxi passenger flow based on a deep learning algorithm. We call the entire framework TBI2Flow. Extensive experiments demonstrate that TBI features has outstanding contribution to passenger flow prediction and TBI2Flow outperforms state-of-the-art methods including time series-based method and other deep learning-based methods on long-term taxi passenger flow prediction.

TBI2Flow: Travel behavioral inertia based long-term taxi passenger flow prediction

Evaluating the impact of scholarly papers plays an important role for addressing recruitment decision, funding allocation and promotion, etc. Yet little is known how actual geographic distance influences the impact of scholarly papers. In this paper, we leverage the law of geographic distance and citations between different institutions to weight quantum Pagerank algorithm for objectively measuring the impact of scholarly papers. The results indicate that the weighted quantum PageRank algorithm can better differentiate the impact of scholarly papers compared to PageRank algorithm.

Xiangjie Kong

Papers

RMGen: A Tri-Layer Vehicular Trajectory Data Generation Model Exploring Urban Region Division and Mobility Pattern

Discovering Transit-Oriented Development Regions of Megacities Using Heterogeneous Urban Data

TBI2Flow: Travel behavioral inertia based long-term taxi passenger flow prediction

Evaluating the Impact of Articles with Geographical Distances between Institutions

A novel adaptive spectrum allocation scheme for multi-channel multi-radio wireless mesh networks