University of Macau

About: University of Macau is a education organization based out in Macao, Macau, China. It is known for research contribution in the topics: Computer science & Population. The organization has 6636 authors who have published 18324 publications receiving 327384 citations. The organization is also known as: UM & UMAC.

Chatbots and conversational agents: A bibliometric analysis

Abstract: Chatbots are replacing some of the jobs that are traditionally performed by human workers, such as online customer service agents and educators. From the initial stage of rule-based chatbots to the era of rapid development in artificial intelligence (AI), the performance of chatbots keeps improving. Chatbots can nowadays “chat” like a human being and they can learn from experience. The purpose of this research is to examine the past research on chatbots (also known as conversational agents) using the quantitative bibliometric analysis. The contribution of this research is to help researchers to identify research gaps for the future research agenda in chatbots. The results of the analysis found a potential research opportunity in chatbots due to the emergence of the deep learning technology. This new technology may change the direction of future research in chatbots. Several recommendations for future research are provided based on the results obtained from our analysis.

Group Sparse Multiview Patch Alignment Framework With View Consistency for Image Classification

Abstract: No single feature can satisfactorily characterize the semantic concepts of an image. Multiview learning aims to unify different kinds of features to produce a consensual and efficient representation. This paper redefines part optimization in the patch alignment framework (PAF) and develops a group sparse multiview patch alignment framework (GSM-PAF). The new part optimization considers not only the complementary properties of different views, but also view consistency. In particular, view consistency models the correlations between all possible combinations of any two kinds of view. In contrast to conventional dimensionality reduction algorithms that perform feature extraction and feature selection independently, GSM-PAF enjoys joint feature extraction and feature selection by exploiting \({l_{2,1}}\) -norm on the projection matrix to achieve row sparsity, which leads to the simultaneous selection of relevant features and learning transformation, and thus makes the algorithm more discriminative. Experiments on two real-world image data sets demonstrate the effectiveness of GSM-PAF for image classification.

Voltage Swell and Overvoltage Compensation With Unidirectional Power Flow Controlled Dynamic Voltage Restorer

Abstract: Voltage swell and overvoltage compensation problems in a diode-bridge rectifier supported transformerless-coupled dynamic voltage restorer (DVR) are discussed. When swell or overvoltage happens, applying conventional in-phase or phase-invariant boosting method causes a rapid rise in dc-link voltage, which may damage the storage capacitors and switching devices, and increase the switching loss. This paper illustrates the inability of the minimum energy injection scheme to solve these problems during unbalanced situations. Therefore, we propose a novel unidirectional power flow control algorithm with DVR maximum compensation limit consideration, which can effectively suppress problems of continuous rising in dc-link voltage. Progressive phase rotating method is applied to prevent compensated load voltage distortion or discontinuity. Appropriate and practical design of the dc storage capacitor's rating is also discussed. Simulation and experimental results for unbalanced voltage swell compensation are given to prove the validity and superior performance of the proposed algorithm.

Electromagnetic Macro Modeling of Propagation in Mobile Wireless Communication: Theory and Experiment

Abstract: The objective of this paper is to illustrate that electromagnetic macro modeling can properly predict the path-loss exponent in mobile cellular wireless communication. This represents the variation of the path loss with distance from the base-station antenna. Specifically, we illustrate that the path-loss exponent in cellular wireless communication is three, preceded by a slow-fading region, and followed by the fringe region, where the path-loss exponent is four. The sizes of these regions are determined by the heights of the base-station transmitting antennas and the receiving antennas. Theoretically, this is illustrated through the analysis of radiation from a vertical electric dipole situated over a horizontal imperfect ground plane, as first considered by Sommerfeld in 1909. To start with, the exact analysis of radiation from the dipole is made using the Sommerfeld formulation. The semi-infinite integrals encountered in this formulation are evaluated using a modified saddle-point method for field points moderate to far distances away from the source point, to predict the appropriate path-loss exponents. The reflection-coefficient method is also derived by applying a saddle-point method to the semi-infinite integrals, and this is shown to not provide the correct path-loss exponent that matches measurements. The various approximations used to evaluate the Sommerfeld integrals are described for different regions. It is also important to note that Sommerfeld's original 1909 paper had no error in sign. However, Sommerfeld overlooked the properties associated with the so-called “surface-wave pole.” Both accurate numerical analyses, along with experimental data, are provided to illustrate the above statements. In addition, Okumura's experimental data, and extensive data taken from seven different base stations in urban environments at two different frequencies, validate the theory. Experimental data revealed that a macro modeling of the environment, using an appropriate electromagnetic analysis, can accurately predict the path-loss exponent for the propagation of radio waves in a cellular wireless communication scenario.

Kuhn–Munkres Parallel Genetic Algorithm for the Set Cover Problem and Its Application to Large-Scale Wireless Sensor Networks

Abstract: Operating mode scheduling is crucial for the lifetime of wireless sensor networks (WSNs). However, the growing scale of networks has made such a scheduling problem more challenging, as existing set cover and evolutionary algorithms become unable to provide satisfactory efficiency due to the curse of dimensionality. In this paper, a Kuhn–Munkres (KM) parallel genetic algorithm is developed to solve the set cover problem and is applied to the lifetime maximization of large-scale WSNs. The proposed algorithm schedules the sensors into a number of disjoint complete cover sets and activates them in batch for energy conservation. It uses a divide-and-conquer strategy of dimensionality reduction, and the polynomial KM algorithm a are hence adopted to splice the feasible solutions obtained in each subarea to enhance the search efficiency substantially. To further improve global efficiency, a redundant-trend sensor schedule strategy was developed. Additionally, we meliorate the evaluation function through penalizing incomplete cover sets, which speeds up convergence. Eight types of experiments are conducted on a distributed platform to test and inform the effectiveness of the proposed algorithm. The results show that it offers promising performance in terms of the convergence rate, solution quality, and success rate.