by Daniel Peck, Research Scientist
This afternoon I spent a bit of time putting together slides for my presentation at AppSec Europe next week on measuring and monitoring malicious (and we'll throw spam in that bucket too) activity on social networks, primarily Twitter and Facebook. I wanted a quick way to show the common threads that go into “bad” tweets, and word cloud came to the rescue. Visualized below are the major threads we've seen through mining the Twitter data over the last month. This is everything that was categorized as either adult, porn, spam/fraud, or malware distribution. No surprise that phrases related to the assassination of Osama Bin Laden dominate the dataset.
It's a little messier than it should be as I have only removed the common English words from this dataset, so the more common Spanish words show up. On the to-do list. For now, I thought you would enjoy.