A New Model of the Public Sentiment Index based on Social Degree of Separation Theory

With the data mining and deep learning developing, more and more data sources are taking into account in the algorithm to get more accurate result. But it is not true that the model with the most data sources is the best one.In this paper, the Cross-Four Degrees Data analysis Model(CFDDM) is proposed to analysis public sentiment index, which is in a new view and combined with social science theory. The theory of the social degree of separation and social attribution are taken into account to analysis. In this paper, the CFDDM use the four degrees of separation theory to analysis the public sentiment, not use the six degrees of separation theory, because the patterns of the information exchange have changed largely than ever,because the developing of internet and new techniques. The four data sources from four different fields are used to analysis the public sentiment index. The weights of different data sources are computed dynamically through probability. The experiment demonstrates that the model is more reasonable and accurate than the models using only one data source.