| | |
Summary: Document Normalization Revisited
Abdur Chowdhury
America Online
Reston, Virginia
chowdhury@ir.iit.edu
M. Catherine McCabe
U.S. Government
Washington D.C.
mcatherm@comcast.net
David Grossman, Ophir Frieder
Illinois Institute of Technology
Chicago, IL 60616
{dagr, ophir} @ ir.iit.edu
Abstract
Cosine Pivoted Document Length Normalization has reached a
point of stability where many researchers indiscriminately apply a
specific value of 0.2 regardless of the collection. Our efforts,
however, demonstrate that applying this specific value without
tuning for the document collection degrades average precision by
as much as 20%.
|
