Using old Spam and Ham Samples to Train Email Filters - Archive ouverte HAL
Communication Dans Un Congrès Année : 2009

Using old Spam and Ham Samples to Train Email Filters

Résumé

Email spam filters are commonly trained on a sample of recent spam and ham (non-spam) messages. We investigate the effect on filter performance of using samples of spam and ham messages sent months before those to be filtered. Our results show that filter performance deteriorates with the overall age of spam and ham samples, but at different rates. Spam and ham samples of different ages may be mixed to advantage, provided temporal cues are elided. The experiments on both corpora show that performance deteriorates faster when using only the body of messages than when using the whole message or headers only.
Fichier principal
Vignette du fichier
ceas09-gvcjm.pdf (146.05 Ko) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04691528 , version 1 (08-09-2024)

Licence

Identifiants

  • HAL Id : hal-04691528 , version 1

Citer

José Márcio Martins da Cruz, Gordon V. Cormack. Using old Spam and Ham Samples to Train Email Filters. Sixth Conference on Email and Anti-Spam - CEAS 2009, Microsoft Research Silicon Valley, Jul 2009, Mountain View, CA, United States. ⟨hal-04691528⟩
6 Consultations
7 Téléchargements

Partager

More