Campus public opinion analysis based on LSTM neural network

In the past decades, public opinion in campus has shown a high incidence trend. Various new media accelerates the diffusion of public opinion. It poses a major challenge to the management of universities. In this paper, a campus public opinion (CPO) analysis model on the basis of long short-term memory (LSTM) is proposed. The new approach, called CPOLSTM, is composed of five operations: data collection, data preprocessing, training and testing, prediction, and decision making. To verify the performance of CPOLSTM, it is compared with two other approaches. Simulation results demonstrate the proposed CPOLSTM can effectively identify the development trend of CPO.