An Agent-based Simulation System for Evolution of Interplay between E-Commerce Vendor and Consumers

Guoyin Jiang, Xingjun Liu, Youtian Wang
2015 International Journal of Hybrid Information Technology  
We use computational experiment method integrated multi-agent, advanced GA (Genetic Algorithm), and complex network to study the interactive evolution of e-commerce vendors and consumers. Key algorithms such as generating algorithm for consumer network, and advanced GA based on CA (Cellular Automata) for service vendor decisions, etc, are designed for the evolution behaviors of both e-commerce vendors and customers. We program the computational experiment system with NetLogo, and a simulation
more » ... periment is given. The experiment system can provide technique support to e-commerce firms in decision making with regards to service providing, consumer behavior management, etc. this paper, we attempt to introduce computational experiment methods to represent the co-evolution scenario. Computational experiment method is a comprehensively integrated method involving computation technology, artificial intelligence, complex system theory and evolution theory [3] ; it can be used to visualize the relation between management scenario and microscopic subjects' behavior via computer. Use of computational method to study system evolutions will help reveal the deep rooted inner workings and microscopic laws of an evolutionary system; so this method is widely used in management and economics when it comes to complex behavior research. Traditional qualitative research can expose concept, framework and observations, while empirical research can analyze influencing factors and behavioral mechanisms on a static basis, that is, the analysis is about an event at a fixed time and a fixed venue; evolutionary game theory can analyze dynamic evolutionary process but with a coarse granularity, and the result is often macroscopic. Agent-based modeling or other computational methods regarding behavior research can study dynamic complex system with a more concrete and more microscopic result [4, 5] . Consumers are sited at a social network, they communicate with other by social media or online tools, for example blogs, Facebook and others. Prior empirical works examine the network structure, the findings indicate that, the www or web is a scale-free network [6, 7] , the network has a small-world feature, where most users are not direct neighbors of one another, but they can be reached from a small number of hops or steps, and the network has a degree distribution following a power law [6, 8] . In this work, we use scale-free network as a surrounding where agent can interact with others. In this paper, we will discuss the implementation of such systems, including main algorithms, and fundamental principles of the interactive evolution between e-commerce service vendor and consumer groups under collaborative e-commerce circumstances.
doi:10.14257/ijhit.2015.8.4.10 fatcat:7xz37gr6ovho3ekydze4mrr2xe