An automatic crisis information recognition model based on BP neural networks

With the popularity of social networks, crisis information has brought huge harm to social stability and people’s lives. The identification of crisis information in social networks plays a crucial role in avoiding the occurrence of crisis events and reducing the harm caused by crisis information. Thus, this paper proposes a crisis information recognition model based on Back Propagation neural network to identify crisis information in social networks. First, we analyze the crisis information and find that the content characteristics, user characteristics and propagation characteristics have a great impact on the identification of crisis information. Thus, we extract the key characteristics of crisis information. Secondly, we label the crisis information. Combined with the extracted crisis information characteristics, we construct a feature tag library for crisis information identification. Then, we use BP neural network to train the feature tag library to obtain the evaluation mechanism of crisis information. When analyzing the suspected crisis information, we use the evaluation mechanism to identify the crisis information. Finally, we evaluate the performance of our recognition model. The experimental results show that the proposed model can effectively identify 97.5 $$\%$$ of crisis information.