May 9, 2006

Microsoft Has Search Result Clustering

MSRA Search Result Clustering is a search engine created by Microsoft that features clustering. Each web page belongs to one or more categories. For Google, the taxonomy contains:

advertising programs(33)
engine(24)
google maps(12)
blog(17)
technology,powerful(9)
support,browser(10)
google answers(5)
google pagerank(10)
inside google(5)
google adwords(5)
tools,hacks(6)
google directory(5)
google book(5)
privacy policy(4)
enable javascript(3)
newsletters,stay(3)
share,pictures(4)
discuss topics(2)
free encyclopedia(3)
google job(6)
others(61)

Between brackets, you can see the number of results for each category.

Clusty, another search engine that uses clustering, has much less categories that are more relevant. For Google, the categories are:

Search engine (41)
Blog (16)
Programs (12)
PageRank (13)
Microsoft (8)
Toolbar (7)
Google Earth (9)
Video (6)
Books (6)
Google Maps (6)

Here's a screenshot from Microsoft's search engine developed at Web Search and Mining Group in MSR, Asia:



Related:
Google semantic Web
Google Cooperation / Google Health

Thank you, Peter.

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.