Last week I attended UMAP 2009 and participated in its Industry Panel organized by Alejandro Jaimes from Telefonica I+D. In the panel, we discussed what are main challenges the industry is facing on personalization. The topic is clearly very hot for both the industry and the academia, and I hope the panel was interesting for the audience. Panelists came from very different types of companies and areas and I am not sure if this was a good thing to have or not. We had Mauro Barbieri from Philips Research (The Netherlands), Susan Dumais from Microsoft Research (USA), Armen Aghasarian from Alcatel-Lucent (Paris, France) and myself from Strands Labs (Barcelona). We discussed important issues for personalization and recommendation such as privacy, modeling, standards, context, and cultural barriers and differences.
There was some controversy about privacy when personalizing on the Internet. I think there is almost no privacy issue on the Internet, I believe this is a topic that we have artificially created (I know I am controversial here!). We are all using credit cards that collect all kinds of sensitive and private data, and we do not really care. This is because of two factors: a) we get a valuable service back that is rewarding enough to us, and b) we trust credit card companies. I think the game is the same when talking about Internet services. I do not mind if Amazon has my reading taste because a) I can buy books online easily and conveniently, and b) I trust Amazon. I believe we (professionals working on the Internet) are not understanding that young people and new generations are actually using the Internet to expose themselves, they want to tell eveybody about everything. MySpace, Facebook, Twitter are all companies about letting people express and expose themselves to the crowd. Of course, they do not care about privacy, they just want to get their lives out there.
I also believe that is very important to understand that data and profiles users are creating on the Web belong to them. Users should have complete control over their data, they own them. Companies hosting these data should be considered just as data hosting and service providers.
Posted by torrens 
