Research on the prediction of user behavior based on neural network

In the rapid development of social networks, the user's behavioral data are becoming more and more concerned by the researchers. Social user data are diversified, including behavior, hobbies, purchasing power, this paper mainly discusses the problem of predicting user behavior in a social network, according to the user's behavior data obtained in the social networking site, Building a prediction model of user behavior, and model evaluation system, combined with neural network algorithm to predict user behavior. By analyzing the behavior data of social users in the past time period, we get the experimental results through neural network model training, indicating that the final judgement level is highly accurate.