InterJournal Complex Systems, 835
Status: Accepted
Manuscript Number: [835]
Submission Date: 2004
Modeling Share Dynamics by Extracting Competition Structure
Author(s): Masahiro Kimura ,Kazumi Saito ,Naonori Ueda

Subject(s): CX.4



The World Wide Web provides a vast information space and continues to grow as a novel important medium of communication. Investigating the Web is becoming an important and challenging research issue, and is also drawing the attention of the research community involved in Complex Systems. From the viewpoints of sociology and economics, the Web can be regarded as a global market in which site maintainers offer information goods to users, and then the number of visitors to a site during a period can become a proxy for that site's success. Namely, we can consider that Web sites offering similar services compete to increase visitors. In this paper, we propose a new method for analyzing fluctuations in the numbers of visitors to Web sites forming a market in terms of competitive dynamics. To achieve this aim, we construct a probabilistic dynamical model using a replicator equation and derive its learning algorithm. This method is implemented for both categorizing the sites into groups of competitors and predicting the future population shares of the sites based on the observed time-series data. We confirmed experimentally, using synthetic data, that the method successfully identifies the competition structure of the true model, and shows better predictive performance than the conventional methods that do not take into account competitive dynamics. We also experimentally demonstrated, using real data of visitors to twenty Web sites offering streaming video contents, that the method suggested a reasonable competition structure (popular entertainment sites, artistic hobby sites and knowledge providing sites) that the conventional method in econophysics failed to find and that it outperformed the conventional methods on predictive performance.

Retrieve Manuscript
Submit referee report/comment

Public Comments: