WebProNews

Time on pages key to determining relevance

It shouldn’t be the links that come in, but the time spent browsing a relevant page, that should help determine where a page ranks for a given query.

Microsoft has a big idea on ranking pages listed in a search engine index for later retrieval. A query seeking information on a topic ought to bring up pages ordered at least in part by how long people spent on such pages in their browsers.

CNET called it a move to one-up search leader Google, which itself only uses PageRank as one of dozens of factors in determining where a page lists in response to a query. Google's dominance of the search market shows their approach is working; Google is even a dictionary-listed verb these days.

A Microsoft research paper shows where the company sees an alternative to a site that in the estimation of many already owns the search market. Through looking at "user behavior data," Microsoft's researchers suggested they have a more reliable way of assessing a particular page's importance.

Microsoft's China-based researchers claim to have this figured out. But it will need data that goes beyond click-throughs to measure adequately.

That brings up an interesting idea: how much access will people, both visitors and site publishers, be willing to give Microsoft to make for a better search experience? Microsoft probably has the ability to utilize its own resources, like the Windows Live Toolbar, to help grab such data.

Not everyone will use a toolbar with their browser, of course, and Microsoft doesn't want to be in the position of trying to push it on people, lest they run afoul of their antitrust minders at the Justice Department.

Google's PageRank succeeds partially due to its invisibility. The typical Internet user simply doesn't see a crawler's work behind the scenes, and computers can accurately count those inbound links in short order. BrowseRank may be the next great advance in search, but if it relies on active participation beyond a link spider's actions, it may not get the critical mass of data it will need to succeed.