Social stability indicator system and its early warning model construction

Abstract

We are now in the midst of a rapidly developing information age, and it is not easy to maintain social stability and overcome social crises in an increasingly competitive and crisis-prone modern society. A social stability mechanism system is an organic whole consisting of multiple types of social stability mechanisms combined in a certain way, which can play a "synergistic" role in stabilising society. How to build a scientific, rational and accurate social stability assessment system and its early warning model is an important step to change the emergency response at the end to maintaining stability at the source, which is a common concern in the world, and is also the main object of this modelling study.

To address question one, we collected data reflecting various aspects of social stability in China from official public databases and statistical yearbooks such as the National Bureau of Statistics of China, the World Bank and the World Trade Organisation, and supplemented and standardised these data with missing values. Taking China as an example, a first-level indicator assessment system was constructed for four dimensions of policy making, social environment, public safety and natural environment, containing a total of 20 sub-dimensions. The rationality of the social stability indicator system is analysed from both quantitative and qualitative perspectives, and the correlations and causal relationships between the four dimensions are explored.

For problem two, the collected data were normalised and standardised, and the prediction targets and feature indicators were selected to build an early warning model of social stability. Based on the processed data, a support vector machine (SVM) model was built, and the warning level was selected as the prediction target and the corresponding sub-dimensions as the feature indicators. The parameters of the SVM model are optimised separately and the model loss is calculated. The parameter corresponding to the smaller loss is the final parameter setting, and the internal parameter settings are adjusted using the grid search method, and the results show that the optimal parameters c=0.25 and g=0.7011.

For question three, Belarus was used as an example to process data related to the country of Belarus by referring to the method used in question one to collect and process data from China. Based on the model already established in question two, the processed data are input into the SVM model as characteristic indicators, thus realising the evaluation of social stability in Belarus. Finally, a grey evaluation system is constructed and the results are obtained using grey correlation analysis and effective suggestions are made for social stability based on the analysis results.

In response to question four, the data relating to the country of Ukraine was approached with reference to the methodology used to collect and process the Chinese data in question one. Firstly, the causes of the colour revolutions leading to regime change in the country of Ukraine were explored. Secondly, the gap between each sub-dimension and the stability warning level was analysed through grey correlation analysis. Finally, the main causes of regime change are identified in relation to the relevant analytical documents and the results of this question.

In response to question five, based on the results of the analysis of the above four questions, and taking into account the relevant literature, practical suggestions are made to prevent colour revolutions and maintain social stability from both the state and individual perspectives.