| Summary: | 碩士 === 國立交通大學 === 資訊科學與工程研究所 === 95 === The thesis presents an agent-based and small-world social network model combining individual hierarchy needs in the social network. Based on various factors, which gives impact on social behavior of individual, we classify these factors into different group and build up the pyramid model with hierarchy needs. These needs include five layers: space-time proximity, complete closure, positive similarity, status continuous, and marginal decline. This model is assumed to be changed continuously and formed a complicated social network with “small-world” and scale-free characteristics. First, the “small-world” has two well-known features: high clustering and low separation. Second, scale-free presents the feature of preferential attachment. That is, the fact the degree distribution of individual abides power-law explains the reason why the deviation of social relationship development exists among individuals. Besides, we also assume that agents act as individuals in the real world and have social relationship by the pyramid of hierarchy needs. From this thesis, we can conclude that our empirical results could reflect social behavior in the real world. Moreover, this thesis also provide sensitivity analysis to evaluate the impact of clustering and separation on the model and tries to the find out connected pattern to understand human propagation behavior for further research references.
|
|---|
