Personal Tracking Devices and Online Identity

Presented at 31C3 (2014), Dec. 27, 2014, 12:45 p.m. (60 minutes)

In the post-NSA world it is important to understand the magnitude of our online activities in order to take informative decisions on our ubiquitous shared lives. Personal Tracking Devices is the result of a two years long study on tracking technologies and the inherent nature of the web and telecommunication networks in general. The study, conducted as part of Ph.D. research in privacy and security at UPC Barcelona Tech, collected a large amount of metadata to raise awareness on the footprints left by users on the web and through mobile apps. Personal tracking devices will visualise the online footprint of a user by looking at their metadata. A hypermedia model of the user footprint would then be introduced in order to better explore it. This model has been called hyperme. Hyperme is a hyperdata model of a user online footprint. The hyperme model links the user identities created across different services and the features associated with them. These features are attributes that compose an identity, such as email, date of birth, place of birth and so on. The hyperme model of the user identity permits the visualisation of the user expressed preferences, the content they have created and who or what can access this content. The model uses context between the user’s various identities and the signals produced, to create links between different objects, obtaining an explorable graph­-like structure. Links between data snippets are creating by exploring the keywords and categories used to describe the entities. These are provided by the user themselves through freeform annotations, particular use of language, location information, timestamps, social relationships and association with other entities such as companies and institutions. Explicit connections are also discovered by associating such keywords with Wikipedia concepts. By exploiting links between articles, it is possible to draw relations between different entities, providing a dictionary to build strong connections between different categories. If the identities created by a single user, and the signals generated are analysed at different levels, it would be possible to discover different subgraphs and sub­hypergraphs between the data object, therefore revealing a complex network of heterogeneous information shared across a number of services and with sets of different parties, being this social relationships or other applications and devices. Each party in fact enjoys a certain level of access to the different documents produced by the user, by the devices used and by the application authorised to access and produce content on their behalf. A hypergraph model therefore allows the possibility to explore the user’s different identities and the corresponding created content at different levels, exposing how different services or relationships contribute to protect or threaten the user privacy.


  • Silvia
    I am a PhD Student at UPC BarcelonaTech, researching in privacy and security, at the Telematics Engineering Department. In the past I was working as Operations Engineer at Google and as Support Engineer for Google Apps. I am particularly interested in privacy and security, and more specifically in analysing dynamics of online social networks and communities, and understanding the underlying structure of crowdsourced information. I have also a keen interest geographical services as I believe this are beginning to be an important part of any information services accessible online.


Similar Presentations: