Better Than You Know Yourself

Over the past 20 years, recommender systems—which predict and suggest merchandise based your browsing history and profile— have become an essential part of the online experience. Vacations, power drills, concert tickets—you name it, these cleverly crafted algorithms connect you with the best deals in the shortest time. The Next Gen of these systems are called ‘context-aware’— machine learning technology that makes precise recommendations based on parameters that range from the weather to whether or not you’re having a good day. You want to buy a shirt or blouse? So: where do you live? what are the trends? how are you feeling these days? –it will all be considered now. It’s a huge jump forward for these sophisticated bits of code, one that can revolutionise the accuracy of online searches, for better or for worse.

“Avoid the Mountain Museum: you’re afraid of heights”
Along with Google and other Internet juggernauts, this technology is being researched by Francesco Ricci and his PhD team at Faculty of Computer Sciences of the Free University of Bolzano/Bozen. “Context-aware technology is very exciting and very difficult… because there are so many conditions and variables that you can use,” says Ricci, a mathematician, who along with PhD candidates Matthias Braunhofer and Mehdi Elahi, has been optimising the technology in several (secret and not-so-secret) products. “’It’s not just bi-dimensional, its multi-dimensional… modeling all these interactions creates models with billions of parameters.” One of the latest innovations the group has developed is ‘South Tyrol Suggests’ (STS), a travel recommender app for Android-based smartphones currently available in beta. The system calculates optimal point-A-to-point-B routes for tourists, along with their best mode of transportation, and advises on what to see along the way. Then it presents the options within the context of the user’s preferences, current weather and traffic conditions. What’s more, with a simple personality scale—degree of extraversion, conscientiousness, openness, and so forth—STS assesses what choices you’d likely make. Why propose a ski tour, after all if you don’t like the unexpected?

No, you don’t want that, you want this.
Oh dear, how much do we want machines to know about us? Ricci points out that cellphones are already broadcasting our location to the world, yet acknowledges that with technology like this, the public may some day be surprised at the reach of Big Data’s omnipotent eye. “When the recommendation becomes so precise to say… “you cannot take the train you normally take because there’s a strike today.” The minute you realise: ‘Oh, he knows that I’m taking the train.’- -this is the critical aspect… I am pretty sure that in the future we will have these kinds of issues and problems because people will realise that the system will know a lot about what they do.” Clearly, a Huxlyan world is not what scientists like Ricci and company have in mind, but the advent of context-awareness will without a doubt transform our relationship with technology.