As well as our impressive new website, today we also took our Big Data & Security research further with an interactive visualization of text message spam in North America.Best viewed in Chrome or Firefox, this builds on our Visualizing Spam model, and reuses Google’s fantastic WebGL Globe. It shows the weekly counts of Blocked Text Message Spam being sent to and from North America during Q4 of 2013, indicating what areas have been targeted and from where.
By cycling through the weeks you can clearly see the general reduction of SMS spam being received in North America during this period. Of particular interest is that it shows you the main areas being targeted by SMS spam in North America, especially southern Florida, which received huge amount of Junk Car Spam compared to the rest of the continent. This visuals show you how much of a problem it really was, and as time goes on how it drops.
From This:
To This:
The visual also shows how the interaction of text message spam between North America and the rest of the world. When 'spam location sent' is selected, you see the locations where spam (that goes to North America) is originated from. While not as sizable as volumes within North America, there are countries which ‘export’ considerable amount of text message spam to the continent, including Guatemala and Ireland. When 'spam location received' is selected, you can see where spam that originated in North America is targeted to. In this case you can see that generally more text message spam leaves North America that enters it, the continent as a whole is a general exporter of text message spam to the rest of the world. You can also see some interesting places where it ends up during this period: such as Iraq and Mongolia.
This is part of our on-going experiments to analyse and gain insight into mobile security data. Feel free to explore and play with the visuals, while the framework was easy to reuse as the code was in place and available, it really came alive once combined with the enriched data we have. The main drawback of the model is that WebGL is not fully supported on some browsers like IE as standard - hopefully this will change in the future, but for now it works great in Chrome or Firefox . As was the case for our earlier visuals, locations within North America are calculated by using NANPA geolocation, - meaning it is not the true location, but the exchange at which the phone is registered. For worldwide locations this is calculated using the country code. This tends to make some peaks outside the US & Canada seem higher/more 'spammy' than those inside, but as the vast majority of text message spam stay within North America it probably evens out. Happy exploring!