Loading...
Paper Search
Current Issue
  15 March 2019, Volume 16 Issue 1 Previous Issue    Next Issue
For Selected: View Abstracts Toggle Thumbnails
An Opinion Dynamics Approach to Public Opinion Reversion with the Guidance of Opinion Leaders   Collect
LIU Qi, XIAO Renbin
Complex Systems and Complexity Science. 2019, 16 (1): 1-13.   DOI: 10.13306/j.1672-3813.2019.01.001
Abstract ( 1117 )     PDF (2214KB) ( 1981 )  
Opinion reversion is an important phenomenon in network emergencies and hot events. In order to explore the internal mechanism and the evolution law of public opinion reversion, we propose an opinion evolution model based on opinion leaders from the perspective of opinion dynamics. The model is applied to simulate the evolution process of public opinion under the guidance of opinion leaders. We take opinion leaders into account, improve the HK model and use social network analysis method to identify the opinion leaders based on scale-free network structure. To verify the simulation results we select the reversal case on sina twitter. The results show that opinion leaders play a key role in the evolution of opinion, which can guide the public opinion to reverse. The results also indicate that the improved model can simulate the evolution process of network opinion reversion.
References | Related Articles | Metrics
Community Tracking Algorithm Based on Similarity of Association Group Evolution   Collect
XU Bing, ZHAO Yawei, XU Yangyuanxiang
Complex Systems and Complexity Science. 2019, 16 (1): 14-25.   DOI: 10.13306/j.1672-3813.2019.01.002
Abstract ( 1028 )     PDF (1763KB) ( 808 )  
In large-scale complex networks, community structure is ubiquitous, and with the change of time, the community in the network is also changing. In order to track the changes of the community and associate the adjacent time groups to form the related groups, this paper proposes a comprehensive weighted evolutionary similarity to measure the similarity of the neighboring time groups. A method of extracting evolutionary path and generating evolutionary sequence by using "multi-part graph" is also proposed. Finally, the experimental results on a bank business data show that the algorithm is more accurate than using a single index similarity judgment.
References | Related Articles | Metrics
Collective Influence Centrality Combining Neighborhood Robustness and Degree Equilibrium   Collect
SONG Jiaxiu, YANG Xiaocui, ZHANG Xihuang
Complex Systems and Complexity Science. 2019, 16 (1): 26-35.   DOI: 10.13306/j.1672-3813.2019.01.003
Abstract ( 1174 )     PDF (1561KB) ( 832 )  
Collective influence (CI) centrality is one of the latest achievements in the measurement of node influences, which is designed based on the locally tree-like network and regards the importance of nodes in the global connection as the representation of their influence. However, it ignores the neighborhood distribution of each node and the difference in the robustness of the local network structure. Therefore, based on CI centrality, the influencing factors in the local network topology such as the neighborhood robustness of the target node, the degree distribution of the l-order neighbors and the connection strength between clusters of the l-order neighborhood are analyzed, defined and quantified. Then, a more universal centrality measurement method called NewCI is proposed to evaluate the influence of nodes. The stability of its overall performance and its better effectiveness and accuracy over CI in node influence measurement are demonstrated by the network invulnerability experiments in six real complex network datasets. Considering the effectiveness, time complexity and execution efficiency, NewCI also has a greater advantage than other commonly used centrality method.
References | Related Articles | Metrics
Quantum-Behaved Discrete Particle Swarm Optimization for Complex Network Clustering   Collect
YANG Zhongbao, CHU Yangjie, HONG Ye, JIANG Dengying
Complex Systems and Complexity Science. 2019, 16 (1): 36-42.   DOI: 10.13306/j.1672-3813.2019.01.004
Abstract ( 891 )     PDF (1614KB) ( 558 )  
Community structure is one of the most important features of complex network. In order to solve the problem of resolution limit of modularity optimization methods, a quantum-behaved discrete particle swarm optimization for complex network clustering is proposed in non-overlapping community detection algorithm (NQD-PSO). The core node and neighborly common nodes are constructed asa motif, which is the initial value of the quantum particle swarm optimization algorithm. At the same time, constructing the motif weighted community clustering function as the adaptive function of the algorithm, while it can use the triangular model to judge the problem of community stability measurement for quantifying the stability of community then the compression factor is adopted to adjust the global and local search model, which makes the algorithm globally converge by combining with quantum particle swarm optimization. Compared with other algorithms, NQD-PSO algorithm uses motif orderly table coding method, and experimental results on both synthetic and real datasets show thatthe NQD-PSO algorithm can mine more high-quality community structures.
References | Related Articles | Metrics
Agent-Based Simulation of Enterprise Entrepreneurship and Innovation Based on DQN   Collect
LI Rui, WANG Zheng
Complex Systems and Complexity Science. 2019, 16 (1): 43-53.   DOI: 10.13306/j.1672-3813.2019.01.005
Abstract ( 845 )     PDF (1651KB) ( 462 )  
Based on ACE (Agent-based Computational Economics), this paper uses an agent-based model to build an economic system model based on enterprise behavior, and tries to solve the dynamic problem and policy problem of combining enterprise innovation with entrepreneurship. In the economic system of enterprise entrepreneurship and innovation constructed in this paper, the agent behavior algorithm, which acts as the enterprise agent, adopts the artificial intelligence DQN algorithm for self-adaptive simulation. The simulation results show that compared with the enterprise agent without self-adaptive behavior, the enterprise agent with self-adaptive behavior is able to make correct business decisions by evaluating the environment and its own state.
References | Related Articles | Metrics
The Enterprise Innovation Evolutionary Games with Time Delays and Switching Topology   Collect
SU Xue, XU Yong, FAN Xujiao
Complex Systems and Complexity Science. 2019, 16 (1): 54-62.   DOI: 10.13306/j.1672-3813.2019.01.006
Abstract ( 1024 )     PDF (964KB) ( 565 )  
Aiming at the analysis of the evolution trend of enterprise innovation time-delay evolutionary game under switching topology, firstly, the cooperative game network is constructed by treating the enterprise as a node and using edges to represent the correlation between enterprises. Then, the model of the evolutionary game with time delays of enterprise cooperative innovation under switching topology is established. Secondly, the revenue of the game, the evolution of the strategic profile and the switching of game networks are formulated separately by semi-tensor product. The algebraic expression of the process of evolutionary game enterprise innovation under switching topology is obtained, and the characteristics of game evolution are analyzed. Thirdly, the control input is designed, the necessary and sufficient conditions for all enterprises to participate in cooperative innovation are obtained. Finally, example is provided to illustrate the theoretical result.
References | Related Articles | Metrics
Why is the “Internet +” Mobile Phone Recycling Mode Ineffective?   ——An Analysis of Evolutionary Game Simulati on Based on mABM   Collect
LI Chunfa, WANG Xuemin, LAI Xixi, XUE Nannan
Complex Systems and Complexity Science. 2019, 16 (1): 63-73.   DOI: 10.13306/j.1672-3813.2019.01.007
Abstract ( 815 )     PDF (3157KB) ( 629 )  
In order to explore the reason why the performance of “Internet +” mobile phones recycling mode is poor, on the basis of clarifying the factors affecting the performance of this recycling mode, an evolutionary game model is established to research strategy selection between consumers and recyclers (platforms). By analyzing its evolutionary stability strategy and stability conditions, the causes of the poor performance of this recycling mode is revealed. The mABM simulation model is used to simulate the evolution process, so as to verify the validity of the conclusions. The research shows that: 1) consumers’ recognition of environmental benefits is the significant factor that restricts the recycling effectiveness; 2) High sensitivity to information leakage losses and transaction convenience costs prevents consumers from adopting “active recycling” strategies, but providing safe and convenient recycling services will reduce the profits of recyclers, resulting in poor recycling performance; 3) The transparency of price and government subsidy distribution are the contradictions between consumers and recyclers.According to the interpretation, some relevant suggestions and strategies are proposed to improve the performance.
References | Related Articles | Metrics
Analysis and Empirical Study of Policy Impact Based on Event Evolution Graph   Collect
SHAN Xiaohong, PANG Shihong, LIU Xiaoyan, YANG Juan
Complex Systems and Complexity Science. 2019, 16 (1): 74-82.   DOI: 10.13306/j.1672-3813.2019.01.008
Abstract ( 1113 )     PDF (1835KB) ( 985 )  
An effective policy impact analysis methods can help the government to understand the impact of the policy on the stakeholders and the market in a timely and accurate manner, which is of great significance for maintaining market stability and social harmony. The paper takes the online review formed after the enactment of the policy as the source data, and based on the LTP, identifying and extracting causal event pairs and succeeding event pairs, using Gephi to construct policy influence event evolution graph, analyzing the impact of policy on stakeholders and related markets, and taking the "317 New Deal" in Beijing as an example to conduct an empirical study. The results show that the policy influence event evolution graph can fully portray the policy's impact on stakeholders and related market.
References | Related Articles | Metrics
Optimal Control of Nonlinear Systems Based on Lebesgue Sampling   Collect
ZHU Mengmeng, SONG Yunzhong
Complex Systems and Complexity Science. 2019, 16 (1): 83-93.   DOI: 10.13306/j.1672-3813.2019.01.009
Abstract ( 1170 )     PDF (1556KB) ( 836 )  
In order to solve the optimal control problem in nonlinear systems, a new event-triggered control strategy based on Lebesgue sampling is proposed based on the performance potential theory. Firstly, according to the optimal control theory, a mathematical model of nonlinear system based on Lebesgue sampling is given. Then, combined with time aggregation method, analytical method and strategy iteration algorithm in Markov decision process, the mathematical model of the constructed mathematical model is solved by Matlab, and the optimal strategy and optimal performance of the system are obtained. Finally, the Lebesgue sampling system is compared with the traditional periodic sampling system. The optimization performance of the two sampling schemes is analyzed in depth, and its advantages and disadvantages of the sampling system are compared. It is concluded that the Lebesgue sampling method can not only improve the system performance, but also solve the "dimensionality disaster" problem of the system. It can reduce the resource consumption of the system to some extent.
References | Related Articles | Metrics