Cell Phone Tracking Reveals Users' Habits

DinkyDogg writes "'New research that makes creative use of sensitive location-tracking data from 100,000 cellphones in Europe suggests that most people can be found in one of just a few locations at any time, and that they do not generally go far from home.' More interesting than their conclusion, however, is how they got their data. 'The researchers said they used the potentially controversial data only after any information that could identify individuals had been scrambled. Even so, they wrote, people's wanderings are so subject to routine that by using the patterns of movement that emerged from the research, "we can obtain the likelihood of finding a user in any location." The researchers were able to obtain the data from a European provider of cellphone service that was obligated to collect the information. By agreement with the company, the researchers did not disclose the country where the provider operates.' Any guesses which European country requires cell phone providers to record where their customers make calls, and then allows them to give that data away without disclosing that they have done so?"

Data has not been anonymized

[hweimer]

Contrary to what the paper suggests, the data has not been anonymized. Proper anonymization means that you cannot derive correlations between the behavior of the individuals, which was the whole point of the paper.

I don't know the exact legal situation in every European country. However, in EU countries this is regulated by the Directive on the protection of personal data [wikipedia.org], which requires for scientific use that safeguards have to be taken to prevent the identification of individuals. For some countries like Germany this means that the data has to be anonymized, although it is a grey area whether pseudonymization is sufficient.

More details on that matter can be found on my blog [quantenblog.net].

Pretty sure it must be the Netherlands

[Idaho]

If you think the USA is bad with regards to telephone taps and the like, try the Netherlands.

Last year, in the Netherlands 25,000 phones where tapped (for different periods of time). These are published numbers (I could link to them but the articles are in dutch only so, well..)

In the USA, the official numbers are somewhere around 2200 phone taps (in 2007).

But that's not all; keep in mind that the USA has over 300 million inhabitants. The Netherlands has only 16 million.

So either the USA government is doing a much better job of keeping even the fact that phones are tapped at all hidden from public scrutiny, or it really is much, much worse here (in this regard, at least).

Potential hint to the country...

[thedrx]

A potential hint as to the featured country might be the name of the author of the project:

âoeSlices of our behavior are preserved in these electronic data sets,â said Albert-LÃszlà BarabÃsi, an author of the project and the director of the Center for Complex Network Research at Northeastern University in Boston. âoeThis is creating huge opportunities for science.â

As if the obvious Hungarian name wasn't enough, his wikipedia entry [wikipedia.org] states he's lived in Hungary and Transylvania. Of course, this might be (and probably is) purely coincidental.

In any case, I, for one, welcome our new PhD vampiric overlords.

Some countries use it to track traffic jams

[Idaho]

Actually some countries use (allegedly) anonymized cell phone data to track traffic jams. This seems to work quite well. At least there have been several experiments and the idea seems promising.

I would consider this a completely legitimate use of the data. However I highly doubt that it is properly anonymized, but that's a different matter.

This could explain why such data was gathered in the first place. If you can still track particular users, it is not anonymized at all however.

Re:Odd conclusion

[Richard W.M. Jones]

Here in Europe, in some countries, cell phone companies offer a service that can reveal a phone's location (with the precision of a fraction of a kilometer/mile) at any given time from any place actually ... Any tracking (except maybe for aid in criminal investigations?) without the owner's consent would be very illegal.

Very definitely this is used in criminal investigations. In the case of the Soham murders [wikipedia.org] back in 2002, one of the victims had a phone which the murderer had turned off. In a public appeal the police said they'd sent a message to the phone, trying to trick the murderer into turning the phone on (which would reveal its location).

In fact this trick didn't work, but mobile phone location data was still crucial. Police plotted all the walking routes [mobilemonday.net] around where the phone was last located just before it was switched off, and from this found the suspect (later, murderer's) house and also disproved his alibi [bbc.co.uk].

Rich.

I Disagree

[FurtiveGlancer]

"Anonymized" may be defined as data that cannot be traced to a named individual. Individuals may still be tracked by other means (arbitrarily assigned number, vice real phone number) to determine patterns without violating individual privacy. So long as they don't specify home addresses, cell numbers or other personally identifiable data, this is valid anonymity.

Of course, this is different from claiming that the data would be used for statistical puroposes only. This study used the data for sample correlations beyond bulk statistical analysis.

Re:Odd conclusion

[Anonymous Coward]

I think you misunderstood. YOU can track "your" phones in Europe. You don't need a court order or a badge. It is usually quite sufficient to send one SMS from the phone that you want to track to permanently enable the "service" for that phone. Afterwards you can use any other phone to request a location update whenever you want. You see, our solution to the energy crisis is to harvest the rotational energy of George Orwell in his grave.

Re:Germany!

[Anonymous Coward]

It is possible that it is Germany this time, but data is being collected in a similar way in Belgium too.

Two points of interest in the Belgian case (the first probably also true for this article):
* You don't have to make a call for them to know your location. Mobile phones are being tracked as long as they're powered on.
* The (anonymized) data are being used for traffic analysis - not just congestion, but also route analysis: how many people reaching Antwerp by a certain highway enter the city, how many visit the harbour, how many just pass by on their way in the direction of Brussels, how many towards Ghent, etc.

Re:Not surprising with 4/gal gas

[Anonymous Coward]

Count yourself lucky to be living in the US, where gas is still cheap.

I filled my tank yesterday, stopped the pump at exactly 100 eur ($157 at the current exchange rate).
Unleaded 95 octane is 1.58 eur per liter here, that's $9.41/gallon.

And you are complaining?

Re:Well, there goes the myth of the EU saner than

[MrMr]

The (putative) sanity of the EU is not really the issue. It appears that the provider and the researchers have violated the EU legislation, and especially "Directive 95/46/EC on the protection of individuals with regard to the processing of personal data and on the free movement of such data" ( http://en.wikipedia.org/wiki/Directive_95/46/EC_on_the_protection_of_personal_data [wikipedia.org] ).

For instance with respect to this article:

Personal data are defined as "any information relating to an identified or identifiable natural person..."

I'm not sure 'anonymizing after billing' as the authors did is sufficient to make the data non-personal (the gist of the article is after all that you can be identified by your stereotypical movements...)

Data may be processed only under the following circumstances (art. 7):

        * when the data subject has given his consent
        * when the processing is necessary for the performance of or the entering into a contract
        * when processing is necessary for compliance with a legal obligation
        * when processing is necessary in order to protect the vital interests of the data subject
        * processing is necessary for the performance of a task carried out in the public interest or in the exercise of official authority vested in the controller or in a third party to whom the data are disclosed
        * processing is necessary for the purposes of the legitimate interests pursued by the controller or by the third party or parties to whom the data are disclosed, except where such interests are overridden by the interests for fundamental rights and freedoms of the data subject

None of those conditions seem to be met...

Re:Well, there goes the myth of the EU saner than

[R2.0]

"at least most western nations outside the US have more choices than tweedle-dee and tweedle-dum"

There is the perception that Europeans have so many more choices when voting, because they have more parties. So, when voting in national elections, they may have a choice between the Democratic Socialists, Socialist Democrats, Labor, Tories, Greens, ad nauseam. And in local elections, the same. So they have 5 different choices of party for 2, perhaps 3 offices.

Contrast this with the US. When I vote this fall, I will have a choice of Democrat or Republican for
- President
- Representative
- Senator (2/3 of the time)
- Governor (1/2 of the time)
- State Representative
- State Senator (1/3 of the time)
- County Executive
- County Council
- Mayor
- City Council
- School Board (often non-partisan, so more than 2 choices)
- Judges (in some jurisdictions)
- And a whole bunch of other minor offices.

So I, and most residents of the US, have well over a dozen choices in our elected representatives. And in the US many voters split their ticket, voting for individuals of different parties during the same election cycle.

While the US 2-party system has its flaws, I would argue that, on the whole, the US system of elections and government is far more democratic than those systems with multiple parties, but few elected posts.

Re:Well, there goes the myth of the EU saner than

[MrMr]

Yes, they are obliged to store non-anonymized data for a maximum period of 18 months, but that crappy legislation only got passed with the explicit provision that it may only be used for specific police inquiries.
The telephone companies are certainly not allowed to do their own data-mining or to hand over that data to varios research groups. In fact, if that has happened here, we may yet see the whole data-retention farce being reversed.
O yeah, period to you too.

Re:Germany!

[Fëanáro]

The map in the paper is from inside some city, and has only cell tower locations.