Unless you explicitly give permission to use your location, interested parties (like, say, advertisers) can only track you with geolocation to within a radius of about 200 kilometers. But researchers in China and the U.S. have figured out a way to get closer--much closer. With a three-stage system using Google Maps, these researchers can, according to New Scientist, get as close as a few hundred meters.
Yong Wang, a computer scientist at the University of Electronic Science and Technology of China in Chengdu, developed a three-stage system to narrow the radius of geolocation without requiring the user's permission. The first stage is the one that's currently used: A packet is sent to the target, and the time it takes to to bounce back is converted into a (very vague) distance. But Wang took it further by realizing that many large organizations, like businesses and schools, usually have their servers in-house, meaning the IP addresses can be tied to a physical location that's easily found. If an IP address is linked to a university, you can just look that university up on Google Maps and have a pretty good idea that the user of the IP address is somewhere nearby.
Wang catalogued about 76,000 such "landmarks" on Google Maps, which leads to the first new stage developed. The new system pings all of the landmarks in that initial 200-km radius, and by analyzing the time that bounce takes, the possible location radius can shrink even further. In this stage, the software might find that ten out of twenty landmarks return similar ping times to the target, and shrink the possible location radius to reflect that result.
After that step, they continue the method until they figure out which landmark is closest to the target. In areas with many targets, like cities, that can lead to unsettlingly accurate tracking--all without the target's permission. The only recourse somebody might have to this method would be to use a proxy, which would effectively confuse the software into returning a null result.
Who would be interested in this? Well, advertisers might not actually care to get this granular with their targeting, although super-local advertisers might be able to target individual neighborhoods. What's more worrying is the possible invasion of privacy from, well, just about anyone, from individuals to small groups to governmental organizations. Is everyone going to have to use a proxy just to avoid any stranger being able to zero in on their laptops? And do the benefits of this kind of tracking outweigh the costs?
Welcome to 1984
"And do the benefits of this kind of tracking outweigh the costs?"
Which benefits are those?
To a user the benefit could be quicker data entry or improved location services on websites without having to have previously visited and acquired a cookie. If you visit a site like Craigslist you could be redirected to your local listings rather than the primary site.
Think of an address entry field that can automatically fill in everything but the street level address of where you currently are, just type it in double check the rest and continue.
My question is what is the cost to an organization to use this?
I would expect nothing aside from purchasing or developing their own algorithm to perform the described process. Afterward the cost would be processing time on their servers to perform the algorithm, however I believe they could offload the function to the client side computers viewing the site.
The cost to the client users is processing time (insignificant) and privacy.
For me as a user I believe the cost is too high, I will hope that our ISP's would be able to mask and obfuscate our IP addresses to limit this without our involvement but who knows what kind of trouble that could cause.
Soon, internet tough guys won't talk so tough.
Most geolocation sites miss my location (if I'm at work) by several thousand miles. I'm physically located in the US, but my employer's main presence is in central Europe.
I'd be curious how accurate it is. :-)
That sort of inaccuracy is annoying when trying to use a site that attempts to enforce "US-only customer" rules on me. I'm in the US, and yet they think otherwise. :-(
300m ≈ 0.10 miles. That's insanely close.
-IMP ;) :)
Anybody else immediately think about why China would love this new system?
@Aldrons Last Hope Your brief statement made me register to comment.
1984 indeed, How can we win with top expert's on their payroll?
"A new method developed by Wang et al. claims to be able to geolocate IP addresses much (they claim 50 times) more accurately (...)"
If you think this doesn't have profound significance for advertising, you've missed the boat on how data mining and analytics work. It doesn't have to be a direct link, where you say "show this ad if you're near this store" (although that is an important case).
The bigger thing is that you gather data over time and build a profile. You can tell a lot about someone just by seeing where they spend their time. Even how much someone moves around is a good indication of what their interests might be.
The survival of the Internet as we know it depends on effective advertising. If you target ads really well, people respond to them. If you target them poorly, they get really annoyed. A small increase in advertising effectiveness has a huge payoff.
The power of building up these profiles is why there are really only a few major ad exchanges. Unless you know quite a bit about your users, you can't target effectively, and people just ignore the ads and get annoyed. We may not love being tracked and profiled, but the alternative is a much less interesting and fun Internet.
Without ads, you will end up paying for every page you view, and lots of fun-but-largely-pointless sites like YouTube would never be sustainable. An advertising-free Internet probably ends up looking a lot like the old AOL: corporate-determined channels, bundled together like a cable package.
@metovek we have to fight for our right to privacy..in the courts and in the streets...peace bro
cant find march :P