Abstract

The Internet is overloading its users with excessive information flows, so that effective content-based filtering becomes crucial in improving user experience and work efficiency. We build Kvasir, a semantic recommendation system, on top of latent semantic analysis and other state-of-art technologies to seamlessly integrate an automated and proactive content provision service into web browsing.

The presentation will focus on the architectural design of Kvasir, and illustrate how to utilise data-parallel paradigm to scale up Kvasir into a practical Internet service. In addition, I will present the solutions to the technical challenges in the actual system implementation, e.g, improving the accuracy of classic random-projection, reducing index size and etc. In the last, I will discuss some promising directions for the future research.

Kvasir was demonstrated in WWW ’15. You can find the techreport, demo videos and part of the source code on the Kvasir website: http://www.cl.cam.ac.uk/~lw525/kvasir/