Prediction of social risk perception on petition in China

Petition attracts more attention because of its unique impact on social life and its increasing trends in China. In this study, we analyzed the causes and classification of petition in terms of social risk perception, and constructed a system of predicting indices by using online big data. First, we reclassified offline petitions in terms of social risk perception, and built online searching indices of certain kinds of petition by using data from “Google trend” and “Baidu index”. Second, we analyzed the predicting effect of social risk perception on online searching indices of petition by Granger causality analysis. Finally, we built an integral predicting model of petition by considering social risk perceptions and online searching indices at the same time. We found that the correlation between offline petitions and Baidu index of petition is more significant than that of Google index. We also found a more significant predicting effect between social risk perception and Baidu index of petition. Moreover, certain kinds of social risk perception such as economy & finance risk perception, have significant predicting effect not only on their corresponding kind of online searching indices of petitions, but also on other relevant kinds of online searching indices of petitions. Therefore, we have demonstrated the possibility of using the correlation among social risk perception indices, online searching indices of petitions and offline petitions to construct online predicting indices of petitions, from which the dominant social contradictions and their relationship in modern China are reflected.