Table of Content

30 September 2020, Volume 38 Issue 5

Previous Issue    Next Issue

Novel Technologies for Intelligent Computing

An Efficient K-Nearest Neighbor Decision Algorithm for Samples with Uncertain Labels

QI Qing, SHEN Zhengfei, CAO Jian, YING Jun, ZHAO Long

2020, 38(5):  659-671.  doi:10.3969/j.issn.0255-8297.2020.05.001

Asbtract ( 403 )   PDF (10944KB) ( 110 )  

References | Related Articles | Metrics

Case-based decision-making is a method to directly classify or predict current cases based on past historical cases. The K-nearest neighbor method is a widely used casebased decision-making model. In the K-nearest neighbor method, historical cases need to be labeled. But in practical applications, the labels themselves have uncertainties. This article discusses the problem of label uncertainty which has been ignored in existing casebased decision-making methods in detail, and setups a label uncertainty model based on Dempster-Shafer evidence theory for improving prediction performance. In addition, in order to improve the operation efficiency, a new boundary tree algorithm by combining the traditional boundary tree algorithm and the label uncertainty is proposed. This paper introduces the function and principle of the boundary tree algorithm, and optimizes the node transfer strategy and decision process of the new boundary tree algorithm. Experimental demonstration shows that the proposed method not only takes the label uncertainty into consideration, but also improves the decision efficiency of the traditional K-nearest neighbor model.

Research on PRBAC Access Control Model in Workflow System

XIONG Tianhong, YU Yang, LOU Dingjun

2020, 38(5):  672-681.  doi:10.3969/j.issn.0255-8297.2020.05.002

Asbtract ( 447 )   PDF (4013KB) ( 136 )  

References | Related Articles | Metrics

Workflow management systems (WFMS) has been widely used in organizational business process management of enterprises and government, and role-based access control (RBAC) model is generally adopted in system tasks for solving the problem of authorization control, and performs good adaptability to the changes of employees; roles or departments. However, with the intensification of competition and the normalization of reform, the organization structures and roles are changing more and more frequently, thus a process system implemented to different organizations will face with much more serious variety of organization structures and roles. The RBAC model causes the task authorization in the business process definition to be heavily organization-dependent, thus the frequent changing of organization will require continuous changing of authorization system, or even worse, lead to its abnormal execution due to the improper process definition. For this problem, this paper proposes a position-role based access control (PRBAC) model, which divides the granularity of roles into organization positions, introduces the concept of business roles which are the only authorization objects, and establishes the corresponding relationship through a mapping layer. The equivalence of PRBAC and RBAC in expressivity is proved, and the granularity and complexity of authorization are analyzed. Through case analysis, we demonstrate that PRBAC model can effectively improve the adaptability and flexibility of WFMS in organizational changes, and realize the decoupling of organization model and business model.

Retrieving Reusable Software by Constructing Functional Descriptions

FU Guangyu, LI Chuanyi, GE Jidong, LUO Bin

2020, 38(5):  682-694.  doi:10.3969/j.issn.0255-8297.2020.05.003

Asbtract ( 457 )   PDF (10112KB) ( 130 )  

References | Related Articles | Metrics

Through software reuse technology, reusing existing software components and modules can effectively reduce the time, labor and costs of new software product development. In software reuse, how to measure and evaluate the reusability of existing software is the first problem to be solved. Although there are a lot of researches assessing the similarities, it is not equal to the reusability. Therefore, this paper defines a set of assessment indexes which is applicable to the reusability of software projects in open source software repository, then designs an algorithm to quickly query reusable software projects based on the basic requirements of the software to be developed, and finally completes the retrieval system of the reusable software project. Experimental results and expert evaluation of the retrieval results verify the efficiency and usability of the method.

Place Bound Algorithm of Petri Net Based on Partial State Space Storage

LU Weihong, DING Zhijun

2020, 38(5):  695-712.  doi:10.3969/j.issn.0255-8297.2020.05.004

Asbtract ( 459 )   PDF (2975KB) ( 125 )  

References | Related Articles | Metrics

Petri net is an important formal modeling tool, and boundedness is one of the important properties of Petri net. In this paper, we focus on this property and propose a new algorithm to solve the place bounds of Petri net without storing all state spaces. The main idea is that in the process of generating state spaces, we can accurately solve the bounds of all places by storing partial states through eliminating some loops of reachability graph and taking advantages of related properties of T-invariant. Experimentally, we compare several other methods with the proposed one in their capability of solving place bounds with the open data set of model checking contest, and results show the effectiveness of the proposed algorithm.

Financial Transaction Data Based Intelligent Fraud Graph Network Detection

SUN Quan, TANG Tao, ZHENG Jianbin, PAN Jing, ZHAO Jintao

2020, 38(5):  713-723.  doi:10.3969/j.issn.0255-8297.2020.05.005

Asbtract ( 544 )   PDF (12309KB) ( 71 )  

References | Related Articles | Metrics

Nowadays, fraud risk has been changing from individual fraud to group fraud, leading to jump rise of financial payment. How to identify and detect the group fraud is becoming a challenge in risk management. To deal with the group fraud transaction, this article builds a transaction graph network based on financial transaction data, and founds a topological graph feature extraction framework and anomaly detection model. Experiment on sample data shows that the proposed model obtains better results comparing with the previous individual feature analysis models, and gives reasonable explanation and evidence for the fraud detection.

Relay-Assisted Offloading Model for Location Privacy Protection under Edge Computing

LIN Wenmin, ZHANG Song, LIU Jiabang

2020, 38(5):  724-741.  doi:10.3969/j.issn.0255-8297.2020.05.006

Asbtract ( 432 )   PDF (11140KB) ( 107 )  

References | Related Articles | Metrics

With the widespread adoption of location-based services, people pay more and more attention to the issue of location privacy protection. The dummy based location privacy protection method could achieve the goal via mixing fake locations with users; real location. However, most of traditional dummy based location privacy protection methods are deployed in remote cloud servers, which brings performance limitation issue such as long delay for users to obtain computation results. To address this problem, we consider migrating the dummy based location privacy protection method from cloud servers to edge servers. Moreover, in view of the upper limit of service capacity and coverage of edge servers, we propose a relay offloading model for location privacy protection. We implement the offloading method and run our method with a real data set. Experimental results verify that our method can reduce the delay for user to obtain computation results while ensuring the effect of user;s location privacy protection.

Multi-task Fuzzy Clustering Based Multi-task TSK Fuzzy System

JIANG Yizhang, HUA Lei, ZHANG Qun, QIAN Pengjiang, XIA Kaijian

2020, 38(5):  742-760.  doi:10.3969/j.issn.0255-8297.2020.05.007

Asbtract ( 774 )   PDF (6981KB) ( 158 )  

References | Related Articles | Metrics

In this paper, a novel modeling approach of multi-task Takagi-Sugeno-Kang (TSK) fuzzy system is presented. Firstly, we propose a new multi-task fuzzy c-means clustering algorithm, which is used to extract the public information between all tasks and the private information of each task effectively. Accordingly, the antecedent parameters of multi task TSK fuzzy system can be constructed with the obtained clustering centers. Secondly, a novel consequent parameters optimization method of the multi-task TSK fuzzy system is proposed based on the multi-task collaborative learning mechanism. Finally, a practical application oriented multi-task TSK fuzzy system is completed based on the two proposed algorithms. Experimental results on several synthetic and real-world datasets demonstrate the validity of the proposed model.

Microservices: Architecture, Communication, and Challenges

DAI Fei, LIU Guozhi, LI Zhang, MO Qi, LI Tong

2020, 38(5):  761-778.  doi:10.3969/j.issn.0255-8297.2020.05.008

Asbtract ( 491 )   PDF (9042KB) ( 83 )  

References | Related Articles | Metrics

Microservices architecture has recently emerged as an architectural style, which is getting a lot of attention in academic community and industry. Microservices architecture utilizes high cohesive microservices and light communication to overcome the problems of poor maintainability and scalability of traditional monolithic systems. In this paper, we apply the systematic mapping study methodology to survey the current state of the art on microservices from following three perspectives: architecture, communication, and challenges. More specifically, we systematically compare monolithic architecture, SOA (service oriented architecture), and Microservices architecture, and then give an overview of the communication between microservices. Finally, we list the technical challenges of Microservices.

Intelligent Computing Offloading for Internet of Vehicles in Edge Computing

MO Ruichao, XU Xiaolong, HE Qiang, LIU Qi, ZHAO Qingzhan

2020, 38(5):  779-791.  doi:10.3969/j.issn.0255-8297.2020.05.009

Asbtract ( 589 )   PDF (2526KB) ( 144 )  

References | Related Articles | Metrics

To meet the requirements of offloading time optimization for computing tasks and load balance optimization for edge devices, an intelligent computing offloading method (ICOM) is proposed in this paper. Initially, a computing offloading model based on the real-world scenario is erected. Besides, the time model of task execution and the load balance model of edge devices are also established. Then, the non-dominant sorting genetic algorithm (NSGA-II) is used to realize the joint optimization of the offloading delay of computing tasks and the load balance of edge devices, so as to find effective computing offloading strategies for computing tasks. Finally, the multi-criteria decision making (MCDM) and the technique for order preference by similarity to an ideal solution (TOPSIS) are utilized to select the optimal computing offloading strategy. Experimental results show that ICOM enables computing tasks to be completed within the expected time, while also ensuring load balance of edge devices.

Clustering by Pruning Paths Based on Shortest Paths from Density Peaks

HU Enxiang, WANG Chunyu, PAN Meiqin

2020, 38(5):  792-802.  doi:10.3969/j.issn.0255-8297.2020.05.010

Asbtract ( 486 )   PDF (3789KB) ( 158 )  

References | Related Articles | Metrics

Clustering is to classify multiple empirical data according to their similarity or proximity based on data labels and properties. For the clustering algorithm based on the density peaks, it mainly focuses on the determination of the clustering center and how to allocate the remaining points. In this paper, according to a trainable clustering algorithm based on shortest paths to density peaks, the clustering center is determined by the density peaks. We propose that using a cutoff threshold and pruning the path graph to improve the algorithm. The remaining points are allocated globally based on the shortest path method. It is proved that the algorithm can significantly improve the efficiency while maintaining the clustering accuracy.

Personality Classification and Conversion Method of Virtual Community Personnel Based on DIKW Graph

LEI Yuxiao, DUAN Yucong

2020, 38(5):  803-824.  doi:10.3969/j.issn.0255-8297.2020.05.011

Asbtract ( 387 )   PDF (3128KB) ( 144 )  

References | Related Articles | Metrics

With the development of social networks, the members of online virtual communities have grown rapidly. In virtual communities, users generally prefer browsing the contents they like, and tend to communicate and cooperate with people with similar or related interests or purposes. The interactive contents between users exist in the form of data, information and knowledge, and generally retain rich “traces” of network users. These traces represent the digital presence of real users. In order to achieve quantitative control of user-generated content in virtual communities based on preferences and interests. This paper proposes to use the DIKW (data, information, knowledge, wisdom) graph to model these typed resources. Combining the user;s DIKW Graph with self-construction theory, users are further classified according to personality index, and the users; psychological needs are also classified. According to the personality index and psychological needs, appropriate personality conversion methods are designed for different users, and the generation of user-preference content is simulated.

Integrated Robust Structured NMF Model for Sample Clustering and Network Analysis

ZHANG Xiaoning, KONG Xiangzhen, LUO Chuanwen, LIU Jinxing

2020, 38(5):  825-842.  doi:10.3969/j.issn.0255-8297.2020.05.012

Asbtract ( 318 )   PDF (6377KB) ( 160 )  

References | Related Articles | Metrics

In order to preserve the homogeneity among data more effectively, this paper proposes an integrated robust structured non-negative matrix factorization (integrated robust structured non-negative matrix factorization, iRSNMF) model with an induced structured term. We verify the effectiveness of this model by applying it to the clustering experiments of cancer samples and the analysis of gene co-expression network. Reasonable biological explanations of related genes and pathways are given based on existing literature. Experimental results show that the iRSNMF method has excellent clustering performance and more-key genes mining ability. The genes and pathways obtained by the iRSNMF model play an important role in cancer pathogenesis, accordingly, providing a new idea for the diagnosis, treatment and prognosis of cancer.