A Multimodal Data Mining Framework for Revealing Common Sources of Spam Images

Abstract

<p class="MsoNormal" style="text-align: left; margin: 0cm 0cm 0pt; layout-grid-mode: char; mso-layout-grid-align: none;" align="left"><span class="text"><span style="font-family: ";Arial";,";sans-serif";; font-size: 9pt;">This paper proposes a multimodal framework that clusters spam images so that ones from the same spam source/cluster are grouped together. By identifying the common sources of spam images, we can provide evidence in tracking spam gangs. For this purpose, text recognition and visual feature extraction are performed. Subsequently, a two-level clustering method is applied where images with visually similar illustrations are first grouped together. Then the clustering result from the first level is further refined using the textual clues (if applicable) contained in spam images. Our experimental results show the effectiveness of the proposed framework.</span></span><span style="font-family: ";Arial";,";sans-serif";; font-size: 9pt;"></span></p

A hybrid spam detection method based on unstructured datasets

“…In particular, larger values of λ imply sparser values of c * . In recent years, sparse representation based classification [3] has achieved promising result in face recognition. In the sparse representation based classification, to classify a new sample, we represent it with training samples from different classes simultaneously.…”

“…Therefore, spam emails can be detected by clustering method based on the similarity of email images. Zhang et al [3] and Chen et al [4] have divided each image into three regions (i.e., text, foreground and background), and performed clustering based on the textual and visual features extracted from the image to detect the spams. Mehta et al [5] represent the image with Gaussian mixture model, and cluster emails through Jensen-Shannon difference to identify spam.…”

A Novel Approach towards Image Spam Classification

“…Section V concludes the paper with a critical discussion. [21] for revealing common sources of spam images. A multimodal framework put forth in [21] clusters spam images so that ones from the same spam source/cluster are grouped together.…”

“…[21] for revealing common sources of spam images. A multimodal framework put forth in [21] clusters spam images so that ones from the same spam source/cluster are grouped together. For this reason, text recognition and visual feature extraction are performed.…”

An Incremental Learning Based Framework for Image Spam Filtering

“…[5,11,12,16,22] are proposed to take advantage of both existing and new anti-spam technologies. With an appropriate combining role, anti-spam frameworks usually achieve higher filtering capacity.…”