Design of Spam Classification System Based on Machine Learning

This article studies the steps of email classification, selects the Naive Bayes algorithm for email classification, uses Support Vector Machine (SVM) for multi label secondary classification, and designs corresponding text files and libraries from the stages of data preparation and preprocessing, garbage labeling, etc. The training model is used to output prediction results and determine whether the email is spam. Using the detailed execution process of the confusion matrix, determine the proportion of the test set for this design, and design the implementation process of primary and secondary classification based on the proportion design.