An Evaluation Model for Public Opinion Formation Based on Entropy

The opinions of individuals change from chaos to order during the formation of public opinion, which is an evolution process of complex system. Many dynamic models have been proposed to explain successfully the mechanism of opinion evolution. Usually, the number of opinion clusters is used to describe the final distribution of individual opinion. However, there is a lack of an unified criterion for evaluation of opinion distribution among individuals. In this paper, the notion of opinion entropy is introduced to quantitative describe the diversity of distribution of opinion. The value of entropy at different situations such as consensus, bifurcation and fragment in the public opinion formation was investigated. From the results, we found that opinion evolution is a process of entropy decreasing, and the opinion entropy is a useful tool in the study of opinion dynamics.