Presentation Layer of Search

Raul Valdes-Perez, president of Vivisimo is quoted in
an article at New Scientist about improvements to the user experience of search interfaces. In particular he is talking about Google News, MSN’s newly announced Newsbot and Vivisimo as yet to be released spontaneous clustering approach.

“[Raul Valdes-Perez] says that the engineering of search and rank algorithms “has gone about as far as it can go”. Now the way to improve the user experience is to work on the next layer of algorithms that determine the presentation of the “search and rank” results.”

“Vivisimo is working on a different approach to the presentation of news search results. Its test news site, which has not yet been revealed to the public, spontaneously clusters links to news articles according to subject.”

In reference to my post yesterday, I don’t completely agree with his statement about the extent of ranking algorithms. In that I believe there is unexploited popularity metadata that should be used as an additional input to existing ranking algorithms. To a small extent this is similar to Amazon’s recommendations, which seems to be one the directions MSN has chosen for its news site.

“This algorithm analyses the other choices of people who have already bought the first book. A news site would instead group articles according to the reading patterns of previous users.”

Although, Vivisimo is certainly on to something in regard to focusing more on the presentation layer of search. Yet, they are not alone in that regard either with existing products such as Antarctica’s Visual Net and new comers like BA-Insight forging through the muck.