Voice and facial recognition technologies to identify barbaric jihadists

Many videos of beheadings by the Islamic State and al Qaeda have appeared online lately. The latest was a video published on November 16, 2014 showing the severed head of 26 year old American Air Force Ranger. Authorities are using voice and facial recognition techniques to identify the man behind this video and other previous videos including the highly publicized killing of the American journalist James Foley.

Linguists are using voice prints of the man on videos and compare all available videos of beheadings. So far they have identified the accent to be of a multicultural man from the London area. They are also saying that the man is deliberately paying more attention to pronunciation of p’s and q’s in order to hide his true identity. Even though “Voiceprinting” technologies are not perfect such as finger print or DNA but many Western countries use it to identify suspects when there is no other evidence.

Pictures on Twitter and Facebook are also valuable resources for investigators who are using photo technologies to identify Jihadi John. Estimated height, weight, eye color and others will also be used. Various computer programs and software are assisting investigators with their work.

How Communities Help to Police Click Fraud on Desktop and Mobile

Click fraud exists, and is fast becoming a widespread problem we should all be prepared to identify it if it happens. Click fraud is a community wide problem. It drives the costs of keywords higher than they should be, creates a false economy around poor performing keywords and costs all of us money. It also means that your market segments aren’t seeing the ads meant for them.

Mobile marketing relies heavily on location-based targeting, and so is susceptible to phony location data. Here, we look at how communities help address this problem by protecting themselves.

How Location Data Works

Locations on a smartphone are pinpointed in a number of ways. Some apps will ask for a user’s location address. Chrome on Android does this, and will use the phone’s GPS unit to pinpoint a user. Most of this location data is accurate up to 100 feet. It’s not easy for sites to spoof this data, but it would require them to send a request to your ad as though they were using a mobile device in the ZIP code you were targeting.

To be clear, this is a rare occurrence. What’s important is being able to understand what is happening to your campaigns. If this helps you rule out location fraud, all the better for you to concentrate on making your campaigns perform better.

The Importance of Location-Based Targeting

Location-specific advertising is supposed to generate users from specific zip codes. When advertisers notice their ads underperforming, it is possible that spoofing is to blame. Ad exchanges test these findings internally to help mitigate the damage of location-based fraud. In this way, ad exchanges preserve a level of trust based on factual reporting and a continued dialogue with the advertisers they work with.

It’s also possible to analyze clusters of devices that seem to be operating out of place. Location data is expected to travel, but not by great distances all that frequently. Searching for clusters helps root out devices that exhibit this rapid-fire movement, and eliminate click fraud.

Final Thoughts

Mobile click fraud is just as real as desktop click fraud, but ad exchanges remain the safest way to purchase traffic because of the safe guards put into place. Because of the testing and vigilance performed by these networks and their communities, ad exchanges are able to find and deal with these problems. The result is a secure network full of quality traffic.

Bio: Ted Dhanik is the CEO and co-founder of engage:BDR, which offers a buy-side platform where advertisers can purchase premium traffic. Ted Dhanik has a background in display advertising, having sold engaging advertising for LowerMyBills.com before founding engage:BDR. Ted Dhanik is based out of Los Angeles, California.