The mathematics of web information retrieval

Friday, February 3, 2006 - 1:25pm - 2:25pm
Ruchira Datta (Google Inc.)
Classical information retrieval was built on the assumption that patient
users are trying to extract information from a clean, trustworthy,
manageable corpus and no adversaries are trying to get in the way. We
can't rely on any of these assumptions in the era of the web. In this
talk we'll see how leveraging mathematics and large-scale systems leads us
to good web information retrieval. Finally we'll describe an interesting
open problem with applications to computing PageRank.