Communications of the ACM
Making Bad Search Results History
Microsoft researcher Jamie Teevan
Photo courtesy of Microsoft
Microsoft researchers are developing personalized search engines that can help users find Web sites by analyzing their past search queries.
Microsoft researcher Jaime Teevan found that more than 25 percent of all searches are aimed at finding information and Web sites that users have already visited. Teevan is developing an algorithm designed to help people navigate to familiar sites without being intrusive. Meanwhile, Microsoft researcher Filip Radlinski and Georgia Tech's Nicolaas Matthijs are developing a system that can perform personalized searches without using any personal information in the search engine.
"To get personalization right, you don't want the user to feel that they are giving away too much information, or to not know what personal information is being collected and where it is going," Radlinski says.
The tool works by tracking which pages a user visits and analyzes their content to create a profile of the user's interests. Then the tool rearranges the top 50 search results to move the more favorable links higher up.
From Technology Review
View Full Article
Abstracts Copyright © 2011 Information Inc. 
, Bethesda, Maryland, USA
No entries found