BSPNN: boosted subspace probabilistic neural network for email security
Artificial Intelligence Review
A survey and experimental evaluation of image spam filtering techniques
Pattern Recognition Letters
A survey of image spamming and filtering techniques
Artificial Intelligence Review
| Hi-index | 0.00 |
In recent years, with the spread of the Internet, the number of spam e-mail has become one of the most serious problems. A recent report reveals that 91% of all e-mail exchanged in 2006 was spam. Using the Bayesian filter is a popular approach to distinguish between spam and legitimate e-mails. It applies the Bayes theory to identify spam. This filter proffers high filtering precision and is capable of detecting spam as per personal preferences. However, the number of image spam, which contains the spam message as an image, has been increasing rapidly. The Bayesian filter is not capable of distinguishing between image spam and legitimate e-mails since it learns from and examines only text data. Therefore, in this study, we propose an anti-image spam technique that uses image information such as file size. This technique can be easily implemented on the existing Bayesian filter. In addition, we report the results of the evaluations of this technique.