More than 140 characters – Tweets: What they revealLe texte complet est seulement disponible en anglais.
Everyone who twitters is aware of one thing: All tweets are publicly visible. Most twitter users are, however, not concerned about their privacy. Why should they? A risk is not automatically created just by the fact that messages are stored permanent and can be publicly viewed. After all, the control about the content remains with the twitter users. This makes sense. But is it true?
Example: Where does your grandma live?
The following example shows that it is possible to detect more in a tweet than the author might know: A Twitter user - let’s call him goofy-567 – twitters a lot and is frequently on the way. His Twitter bio does not contain any information about his home or his job. All in all, the content of information of his messages is harmless and doesn’t allow many conclusions about the author. In general, he tells about what he is eating or drinking at the moment or which event he will attend. Despite this, it is possible to find out by means of his tweets how his girl-friend is called and where she lives, where goofy-567 works and what he does in his spare time. It is even possible to find out how the house of his grandma looks like who does not use Twitter herself. How is this possible?
Geotagging makes it possible
Via Twitter, it is not only pure content that accesses the Internet but also position data. This is always the case when Twitter users locate their tweets. Twitter offers a location determination. In the tab „Security & Data Protection" in the settings, the user can activate this determination by placing a tick mark at "twitter position". This position information is located via the IP address when the browser is set-up accordingly. It is, however, quite imprecise. Geotags that are exact down to longitude and latitude are possible with the recent versions of the iOS and Android Twitter apps, with third party applications, websites or the API of smartphones and tablets. You only have to click on the button „Share precise location". Thus, the localization of single tweets is possible, too. The user does not have to activate a general release for geotagging. As soon as Twitter users switch to automatic geotagging all tweets are tagged.
The power of connection: 1 + 1 = 1000
Independent of the fact that geo-data are transmitted knowingly or unknowingly, for all tweets or only for a few ones – these data can always help third parties to more information about the authors of tweets and posts than the authors be aware of. It is quite simple to make these implicitly transmitted data visible. You just have to connect two freely available, independent internet services such as Twitter including geo-data and Google Maps. This connection results in a complete new data level containing information about the Twitter user that has not been entered in this way. You can, for example, find out where a Twitter user lives just by having a look from which place most tweets are posted. In the second step, you can take a look at the content of the tweets to check your conclusion. If the tweets presumed to be posted from the home of the twitter user are about the current TV program or about his breakfast, your conclusion is confirmed. You can find out more according to the same pattern: When and where to did the user make a journey? How long did he stay? Which placed did he visit and when? And – in case he has twittered such content – what he did there, whom he met, about which subjects they talked. Thus tweets that only describe harmless every day events can be used to create a motion profile, a sociogram and a field analysis.
With the Google services „Earth" and „Street View" you can even find out much more about the determined addresses: the state of the home (at the moment of the "Street view" picture), the neighborhood and the quarter somebody lives in - all this allows conclusions about the financial situation and the social class of a person. Whether these conclusions are correct or not is of secondary importance. In most cases, the impression will be considered as truth.
The opposite approach of gaining information is possible, too. It is easy to find out about which subjects are preferably twittered at a certain place. In doing so, you can also get information about people who are not registered at Twitter or are not active themselves. A tweet from goofy-567: „Stuck in a traffic jam. I hope grandma keeps the apple pie warm." might be harmless. If this tweet is, however, followed by a "finally arrived" transmitting the precise position the address of the grandmother can be easily identified – including name, phone number, etc. This is, of course, not only applicable for the grandmother but for all other people, too, who are not registered at Twitter, Facebook and others: There are already millions of unsaved personal data freely available in the internet. This is a fact everybody should be aware of – not only those who actively use social media services.
Very often, social media users transmit information that might be used to detect personal data by connecting freely available internet services. This is possible by so called geo-tags (position information) that allow the creation of motion profiles, field analyses and sociograms. The meta data are even available for persons who are not registered at social networks or not active.