Semantic analysis for spam filtering

Many different techniques have been employed to analyze spam emails. The paper explores two main semantic methods: Bayesian algorithms and Support Vector Machine (SVM). More recent spam filters are introduced in the paper. They all utilize semantic analysis information to determine whether a message is spam.