Consider8

The Consider8 project will provide methods and algorithms for refining mobile communication traffic data into patterns of human mobility. The aim is to provide results that will assist in meeting existing market needs smarter and faster than current state-of-the-art, while adding much greater sensitivity to ethical considerations such as privacy.

The most pressing concerns for the project is to strike a balance between the privacy of mobile communication subscribers (individuals as well as groups) and the exploitation of traffic and transaction data. One of the main project activities will therefore be to chart the range of privacy issues that this data use may expose, while other activities will focus on methods and algorithms for creating patterns of human mobility in ways that make it impossible to identify individuals.

Among other issues, the project will investigate:

  • the role of anonymization vs. pseudonymization,
  • the applicability of informed consent and revenue sharing, and
  • technical issues related to scalability and robustness of information collection and processing.

The project is a collaboration between SICS and Ericsson Research. SICS participation is fully funded by VINNOVA.

For more information about the project, please contact Markus Bylund.

Publications
Number of items: 2.

Görnerup, Olof (2012) Scalable Mining of Common Routes in Mobile Communication Network Traffic Data. In: Tenth International Conference on Pervasive Computing, 18-22 June 2012, Newcastle, UK. (In Press)

Sanches, Pedro and Bylund, Markus (2011) Experiencing mobility data. In: Workshop on Values in Design - Building Bridges between RE, HCI & Ethics held in conjunction with INTERACT 2011 13th IFIP TC13 Conference on Human-Computer Interaction.. (In Press)

This list was generated on Mon Apr 21 12:58:20 2014 CEST.
Research Issues

There are several pressing research questions related to the urge to exploit information generated by the use of IT.

First, there is a range of ethical issues that must be addressed. One does not have to be too imaginative to come up with a number of use cases that would clearly violate the privacy and autonomy of individuals. But the stories of Phorm and NebuAd make imagination unnecessary. On the other hand, no one can claim complete and exclusive right to all personal information related to her. This would require total isolation. Interacting with other people and society inevitably demands a compromise between private and public. Few people get upset by the fact that most stores have optic counters in their entrances for keeping track of the number of customers. Likewise, few drivers seem to avoid road segments with sensors that not only count the number of vehicles passing, but also classify them and measure their speed.

There is clearly a need to understand where to draw the line. What type of informational byproducts can be used without the risk of violating someone? For what uses can the information be exploited? These issues are intimately related to informed consent, but also issues of creating realistic and robust expectations on IT use.

In addition, there is a whole range of research issues related to these basic questions of secondary use of informational byproducts. For example, what is the role of anonymization? Would pseudonymization make a difference (at least Phorm was hoping that it would), or perhaps information aggregation?

Second, in the cases where refinement of informational byproducts can be considered ethically sound, we are facing a number of related challenges. For example, most use cases will be targeting an abundance of information, thus creating challenges related to scalability. There are also a number of challenges related to what type of predictions can be made of informational byproducts, and their accuracy.

Background

Curiosity over other people’s whereabouts, doings, and opinions has always been strong, for many different reasons. Merchants and marketers are interested in understanding customers and their preferences. Politicians are interested in people’s opinions in a whole range of issues. And most people are interested in what just about everyone else do, say, and think–just look at fashion and trends. The methods for collecting this type of information are many, ranging from simply throwing a glimpse at the square outside the shop to assess the customer base of the day to formal polls and statistical analysis.

With broad adoption of information technology throughout society, there is a whole range of new information that can be used to gain an understanding of people’s behavior. This information can typically be classified as byproducts of some other activity, including the use of cash and credit cards, automatic toll gates, cellular telephony, wireless networking, and just about everything related to the Internet.

However, how to ethically exploit this information is not straight forward. This is clearly illustrated by the failures of Phorm in the UK, and NebuAd in the US. Both companies were mining the Internet habits of individuals by means of deep packet inspection of Internet traffic in order to target direct advertising. Both companies have received massive critique from both the public and governments.