Simulation of Encouragement Methods for SNS based on User Behavior Model

There are a variety of user-limited SNSs as well as large scale SNSs. Since user-limited SNSs are not sufficiently utilized, finding a method to increase their use is urgent. The purpose of this study is to find an effective method to encourage participants. However, from the viewpoints of time and cost, improving services is unrealistic. Therefore we use agent-based simulation to find methods for activating inactive SNSs. To design agents based on user behaviors, we analyzed users in real SNSs. By applying a clustering method, we classified them into six groups and designed six types of agents from the groups. By applying a clustering method, we classified SNSs into four groups based on their user compositions and compared the effects of the encouragement methods by simulations.We clarified that effective utilization promotion methods are different from types of SNSs, which are classified by the constitution of users.

Date: 
Tuesday, 21 September, 2010 - 12:30 to 14:00
Presenter(s): 
F. Toriumi