Evaluation of Spam Detection and Prevention Frameworks for Email and Image Spam - A State of Art

1. Evaluation of Spam Detection and Prevention Frameworks for Email and Image Spam - A State of Art Pedram Hayati, Vidyasagar Potdar Digital Ecosystems and Business Intelligence (DEBI) Institute Curtin University of Technology http://debii.curtin.edu.au/~pedram/ http://debii.curtin.edu.au/~vidy/

2. Outline • INTRODUCTION • SPAMMING MOTIVATIONS • ANTI-SPAM STRATEGIES • EMAIL SPAM • EMAIL SPAM METHODS • EMAIL SPAM SURVEY • IMAGE SPAM • IMAGE SPAM METHODS • IMAGE SPAM SURVER • CONCLOUSION

3. INTRODUCTION • Rapid adoption of the Internet • The ease with which content can be generated and published has also made it easier to create spam. • Spam can be simply stated as information which does not add value to the web user • E.g. inappropriate, unsolicited, repeated and irrelevant content in email messages, search results, blogs, forums, social communities and product reviews. • the aim of this paper is to survey the current literature in the field of anti-spam with focus on specific anti-spam techniques used in email spam and image spam.

4. SPAMMING MOTIVATIONS • Revenue Generation • publishing advertisements on websites. Google AdSense™ • Higher Search Engine Ranking • incorporate search engine optimization techniques to get their website a higher rank in search results • Promoting Products and Services • spammers are paid by companies to promote their products or services • Stealing Information • setup hidden programs in user’s computers to gain back door entry • Phishing • to steal sensitive information (such as credit card numbers, password, etc.)

5. ANTI-SPAM STRATEGIES • Spam Detection Strategy • Try to identify the likelihood of spam in a system either automatically or manually • Spam Prevention Strategy • Deal with the problem of spam in different way. In this strategy, system designers create challenges for the spammers and make spamming a difficult task

6. ANTI-SPAM STRATEGIES Spam Detection Strategy Spam Prevention Strategy

7. EMAIL SPAM • Email spam refers to sending irrelevant, inappropriate and unsolicited email messages to numerous people • Low entrance barrier and low cost of sending emails, which makes it one of the most popular forms of spam • The purpose of email spam is advertising, promotion, and spreading backdoors or malicious programs

8. EMAIL SPAM METHODS

9. EMAIL SPAM SURVEY • Majority of methods are detection based. • Majority is not suitable for non-English language. • Behaviour-based methods can not detect spam on the fly and spammers can manipulate their behavior to game the system easily.

10. IMAGE SPAM • Spammers put their spam content into the images. • They embed text such as advertisement text in the images and attach these images to emails. • Anti-spam filters that analyse content of email cannot detect spam text in images

11. IMAGE SPAM METHODS

12. IMAGE SPAM SURVEY • OSR-based methods are vulnerable to content obscuring tactics and reasonably slow. • Meta-based methods can be easily exploit by changing image file attributes. • Template-based methods are vulnerable to real world images.

13. CONCLOUSION • Behaviour-based methods need time to create user profile can not detect spam on the fly. • Needed methods to detect spam on non-English content. • OSR-based methods are slow and vulnerable. • Template-based methods are vulnerable to real world images

14. Thanks! Any Questions?