In the digital age, data is ubiquitous. The increasing ease with which every online interaction can be stored, compared and analyzed has transformed a wide swath of business and led to the formation of new ones. The same principles apply to the scientific world, where fishing correlations from the sea of millions or billions of data points promises to reveal new insights about the genetics of disease or the properties of never-before-seen molecules.

While a future where algorithms allow taxis to know where riders will be, online stores to know what new items shoppers will want, and medical organizations to know where an outbreak will next strike sounds ideal, the other side of this coin is that not everyone wants the world to know what these algorithms might reveal.

The stakes are even higher in the context of national security. Network analysis techniques that comb through an individual’s social connections are particularly suited to the threats of the 21st century, but are also particularly pernicious when it comes to issues of surveillance and privacy.

Researchers at Penn’s Warren Center for Network and Data Sciences are exploring how the algorithms that conduct this kind of analysis can be designed to guarantee certain privacy protections.

(via Balancing Privacy and Security in Network Analysis | University of Pennsylvania)