ICS-SVM: A user retweet prediction method for hot topics based on improved SVM

Abstract

In social networks, many complex factors affect the prediction of user forwarding behavior. This paper proposes an improved SVM prediction method for user forwarding behavior of hot topics to improve prediction accuracy. Firstly, we consider that the improved Cuckoo Search algorithm can select the optimal penalty parameters and kernel function parameters to optimize the SVM and thus predict the user's forwarding behavior. Secondly, this paper considers the factors that affect the user forwarding behavior comprehensively from the user's own factors and external factors. Finally, based on the characteristics of the user's forwarding behavior changing over time, the time-slicing method is used to predict the trend of hot topics. Experiments show that the method can accurately predict the user's forwarding behavior and can sense the trend of hot topics.