24 Aug 2004
CS 5244: Indexing / Classification
53
MeURLin
ˇClassification of URLs to the
Open Directory Project
ˇ
ˇhttp://www.onlineshawnee.com/stories/072901/ent shelton.shtml
ˇ
ˇDoesn’t require webpage, just address
ˇAbout 1/2 - 1/3  as accurate as full words approaches
ˇUses scalable segmentation and expansion techniques