Comparative Study on SMS Spam Recognition Models Based on Neural Networks

With the rapid development of the mobile Internet, SMS spam has evolved into a key constraint affecting user security and network security.Accordingly, a model based on convolutional neural networks and a model based on BERT are designed for SMS classification. Then, the model is trained by the training dataset, and the neural network parameters are adjusted to improve the classification accuracy. Finally, the results are compared with traditional machine learning methods, and the model is evaluated and optimized by the multi-fold cross-validation method, which significantly improves the model performance. The experimental results show that BERT model performs better in recognition accuracy and practicality, which contributes effective preventive measures for users and operators.