부산대학교 도서관

This paper focuses on the recognition algorithm of spam messages, and collects and arranges the data through the research on the current situation of spam messages at home and abroad. At the same time, businesses seek benefits by sending spam short interest such as advertising and fraud, which seriously endangers the information security and normal life of SMS users. Based on the text content of SMS, this paper establishes a recognition model to accurately identify spam SMS, so as to solve the problem of spam SMS filtering. Using python programming, Gaussian naive Bayes is used to realize text classification, and a high classification accuracy is achieved. The experiment shows that the recognition rate of this algorithm is 93.72%. Experiments show that the improved Bayesian method can effectively identify spam messages.