Scalable personalization infrastructures

The ever-growing amount of data available on the Internetcalls for personalization. Yet, the most effective personalizationschemes, such as those based on collaborativefiltering (CF), are notoriously resource greedy.We argue that scalable infrastructures should rely on P2P designto scale to that increasing number of users, data and dynamics.
I will present a novel scalable k-nearest neighbor protocol, which P2P flavor providesscalability by design. This protocol provides each user withan implicit social network composed of users with similar tastes in a given application.

This protocol has been instanciated in various settings:
(1) A P2P system, WhatsUp, a collaborative filtering system for disseminatingnews items in a large-scale dynamic setting with no central authority;
(2) A hybrid recommendation infrastucture HyRec, an online cost-effective scalablesystem for CF personalization, offloading CPU-intensive recommendationtasks to front-end client browsers, while retaining storageand orchestration tasks within back-end servers;
(3) A cloud-based centralized recommendation engine.

Experiment show that our solution outperforms alternatives with respect to costwhile maintening the quality of personalization.