It's also a great result to have at our fingertips, published and thought through by some brilliant minds at Google so we don't have to.
On Tue, Jun 10, 2008 at 3:54 PM, Sebastian Benthall <[EMAIL PROTECTED]> wrote: > That sounds like a pretty good result for this kind of research. > > > On Tue, Jun 10, 2008 at 3:46 PM, Joshua Bronson <[EMAIL PROTECTED]> > wrote: > >> Good find. >> >> On Mon, Jun 9, 2008 at 3:02 PM, Randall Leeds <[EMAIL PROTECTED]> >> wrote: >> >>> Hopefully useful. Reading through it now and will report back. >>> >>> http://www2007.org/paper570.php >>> >>> Enjoy. >> >> >> So I kept getting this dirty feeling throughout the article that it was a >> lot of sketchy work ("let's see what happens if we throw this in!") for only >> a little result. Like, three or four brilliant minds at Google spent >> considerable time and effort to get a 38% improvement in click-through rates >> on recommended articles over the baseline popularity algorithm. Is that >> actually worth it, or am i just being an (overly idealistic) n00b? >> > >
