Mechanisms to detect Spam

• Email header and structure analysis
• Content analysis through weighted word lists
• Self-learning Bayes network
• Checksum server
• Text analysis using statistical methods

Example

Spam Assassin:

1. Check Header and body text, evaluate and extract features
2. Compare how many times a certain feature has been found in spam mails
3. Create a probability for that feature
4. Combine all probabilities to get spam mail probability

→ If user marks a spam, self-learn

Decide if it’s a spam:

Anti-Spam Protocols

SPF (Sender Policy Framework)

DNS Text query for mail address in “MAIL FROM”

Possible Responses:

• No published SPF record