Anomaly-based Spam Filtering - SECRYPT 2011
0 likes412 views
The document discusses anomaly detection approaches for identifying spam emails. It compares the performance of two spam detection tools, SpamAssassin and Ling Spam, using both Manhattan and Euclidean distance thresholds to classify emails as spam or not spam. Results show that SpamAssassin achieved the highest F-measure of 94.62% using minimum Manhattan distance, while Ling Spam achieved the highest F-measure of 92.20% using mean Euclidean distance. Anomaly detection approaches may help overcome the large volume of unclassified spam emails.
Anomaly-based Spam Filtering - SECRYPT 2011
- 2. WHAT YOU GOT, THEN? SPAM, EGG, SPAM, SPAM, BACON AND SPAM. SPAM, SPAM, SPAM, BAKED BEANS AND SPAM. ANYTHING WITHOUT SPAM? I DON’T LIKE SPAM!! UGH!
- 3. Meet the real SPiced hAM
- 4. Monty Python’s Flying Circus
- 5. Something that repeats and repeats until being annoying
- 6. It is a real problem for Information Security
- 7. Billions of daily losses in productivity
- 11. We must fight
- 12. Anti-spam methods Pre-sending New protocols Post-sending Increase sending costs Increase risks for spammers E-mail sender E-mail content E-mail content
- 14. A significant labelling work is needed
- 15. A significant labelling work is needed
- 19. YES
- 21. no interest this SpamAssassin word has this has Ling Spam no interest word SpamAssassin Ling Spamt1 t2 t3 D1 D2 D10 D3 D9 D4 D7 D8 D5 D11 D6
- 22. ? ? Anomaly detection d d> threshold? > threshold?
- 25. Minimum distance Maximum distance Mean distance
- 28. Anomaly detection d d < threshold > threshold
- 30. Minimum distance Maximum distance Mean distance Manhattan distance Euclidean distance 10 thresholds
- 31. Results
- 32. SpamAssassin Manhattan Euclidean Prec. Rec. F-Meas. Prec. Rec. F-Meas. Mean 91.03% 92.85% 91.93% 76.14% 97.77% 85.61% Maximum 69.61% 99.89% 82.05% 72.99% 97.66% 83.54% Minimum 95.40% 93.86% 94.62% 92.10% 94.00% 93.04%
- 33. SpamAssassin Manhattan Euclidean Prec. Rec. F-Meas. Prec. Rec. F-Meas. Mean 91.03% 92.85% 91.93% 76.14% 97.77% 85.61% Maximum 69.61% 99.89% 82.05% 72.99% 97.66% 83.54% Minimum 95.40% 93.86% 94.62% 92.10% 94.00% 93.04%
- 34. Ling Spam Manhattan Euclidean Prec. Rec. F-Meas. Prec. Rec. F-Meas. Mean 79.18% 73.54% 76.26% 92.82% 91.58% 92.20% Maximum 76.23% 74.29% 75.25% 85.95% 79.29% 82.49% Minimum 65.82% 74.38% 69.84% 87.51% 93.13% 90.23%
- 35. Ling Spam Manhattan Euclidean Prec. Rec. F-Meas. Prec. Rec. F-Meas. Mean 79.18% 73.54% 76.26% 92.82% 91.58% 92.20% Maximum 76.23% 74.29% 75.25% 85.95% 79.29% 82.49% Minimum 65.82% 74.38% 69.84% 87.51% 93.13% 90.23%
- 36. Suitable to overcome the amount of unclassified spam e-mails
- 39. Will we see the END of spam?
- 40. 95%
- 41. “Solution to spam” Cut their billing systems?
- 45. References 1. Monty Python – Spam: http://www.youtube.com/watch?v=anwy2MPT5RE 2. Spam wall by freezelight: http://www.flickr.com/photos/63056612@N00/155554663/ 3. monty python flying circus by the_d8_show: http://www.flickr.com/photos/8056839@N04/478599790/ 4. Dollars: http://vegasgravy.com/News-detail/two-women- caught-for-transporting-drug-money-from-vegas/dollars/ 5. Day 97: Infected by dustywrath: http://www.flickr.com/photos/10921499@N07/2187318683 6. my bank sucks by B Rosen: http://www.flickr.com/photos/rosengrant/3537904106/ 7. Feet on table: http://bisystembuilders.com/wp- content/uploads/2010/02/shutterstock_feet-on-table.jpg 8. Buried on bills: http://getupkids.net/wp- content/uploads/2013/06/debt_piling.jpg 9. Kill spam: http://www.email-marketing-wizard.com/wp- content/uploads/2010/03/spammer.jpg
Editor's Notes
- #2: [click]
- #6: After that, the internet community associated the term spam to repeated and annoying messages.So, next time you receive an e-mail message of a pharmaceutical company offering you viagra, and you think about the mother or father of spam, think about Hormel Foods.[click]
- #8: Causing billion dollar losses daily around the world for lack of productivity[click]
- #9: Infecting millions of computers[click]
- #10: And affecting users who lose money as result of stolen credentials[click]
- #11: For each 5 mails sent,[click]4 are spam.[click]
- #45: and fight this serious problem everyway we can.
- #46: and fight this serious problem everyway we can.