Agent-Based Creation and Simulation of Artificial Social Networks and the Analysis of Their Properties

摘要

Social relations and social networks are a vital part of our society. However, it's difficult to find out what a network might look like under slightly different circumstances. This is where artificial models and simulation can help. With the correct model, a wide range of networks can be created and their properties investigated. Such networks can be used to test and validate various hypotheses. In this article, the authors propose an agent-based model that allows for the building of networks that have properties similar to those observed in the real world. They've tested this model with a significant range of parameters and show how these parameters impact the resultant network's properties. They've also found that the currently used method for estimating the cut-off value for power law applicability is very sensitive to small changes in the evaluated distribution.