Public opinion hot topic discovery based on improved bacterial foraging algorithm

Abstract

In order to find the hot topic from public opinion and solve the problem that the topic mining accuracy is not high enough, the topic is disturbed by unrelated links and the topic cannot be focused on a point, the swarm intelligence algorithm is applied to topic mining, and an improved bacterial foraging algorithm is proposed. The algorithm improves the chemotaxis, copy and migration operations in the original algorithm, which makes up for the shortcomings of the original algorithm, and greatly improves the accuracy and convergence. The correlation value of web page is taken as the measurement of web page heat, and the heat evaluation model is established. The improved bacterial foraging algorithm is used to cluster the web pages. The results of experiments and other algorithms show that the algorithm has better clustering effect.