What percentage of Twitter photos are geotagged

I know where you live: Twitter photos reveal more than we'd like

The website ICanStalkU.com searches Twitter messages for images around the clock and publishes the location of the respective user in almost real time. This webstalking is made possible by the geodata that smartphones nowadays usually automatically integrate into the photos. Name, avatar and whereabouts on Google Maps are clearly arranged. Additionally, the page displays the original message and the tell-tale picture. In addition, the user whose information was tapped is notified via Twitter. This shows that, like the Please rob me offer, it was not developed for the purpose of unnoticed spying on strangers.

With their project, the founders of the site want to draw attention to the fact that many of the lightly published photos say even more than is apparent at first glance. That's why the subtitle of the page is “What People Are Really Saying With Their Tweets”. The basic equipment of many telephones now includes a GPS function that can determine the whereabouts of the owner. This is often desired, after all, this function forms the basis of location-based services such as Foursquare and Gowalla. At the same time, the captured images are invisibly provided with geodata by default. Some users don't care at all, explains security advisor Larry Pesce from ICanStalkYou.com. Others, however, are completely unfamiliar with the additional function of the telephone camera and are accordingly shocked when they are informed about it.

If you look at the location information, the news and the pictures together, you can easily find out "where people live, what things they have in their homes and when they will be leaving," warns Robin Sommer of the University of Berkeley in a recently published study ( PDF). With a little programming knowledge, it is possible, for example, to search the service for holiday announcements in a certain district. “Anyone sixteen with basic programming skills can do that,” adds his colleague Gerald Friedland.

To illustrate the practical dangers, Kate Murphy in the New York Times chooses the example of the presenter Adam Savage. He had shown the fans of his popular science program "MythBusters" on the short message service a photo of his car and wrote: "Now let's go to work." That made it pretty easy for potential burglars. You would now know exactly where he lives and that he is not at home, so Murphy. He also revealed the address of his private house to potential stalkers; he could have avoided this by switching off the geodata function on his phone.

Free LinkedIn guide!

Register now for our weekly BT compact newsletter. As a thank you we give you ours LinkedIn guide.


For the average user, however, this shouldn't be that easy. Murphy criticizes the cumbersome deactivation: "To switch off the geotag function you have to search through various menu levels until you find the" Location "setting and select" Off "or" Do not allow "." Here, too, ICanStalkYou promises .com Remedy: on the page there are step-by-step instructions for switching off the corresponding function.

Some online photo services have already responded to phone manufacturers' failures. TwitPic, for example, offers to save photos with or without location data when uploading photos (see screenshot). Flickr, too, is now preventing the geospatial data from being published without the consent of the user. However, the problem is not limited to image services alone. According to the study by Sommer and Friedland, similar dangers emanate from careless publications on Youtube and Craiglist. After the function has been switched off, the problem of geodata in the photos that have already been published remains of course.

(Nils Baer)