SP500207
NIST Special Publication 500-207: The First Text REtrieval Conference (TREC-1)
CLARIT TREC Design, Experiments, and Results
chapter
D. Evans
R. Lefferts
G. Grefenstette
S. Handerson
W. Hersh
A. Archbold
National Institute of Standards and Technology
Donna K. Harman
Step: Input: Process: Output:
4d Rel-DocsT0[OCRerr] IThesaurusi Pseudo-ThesT.[OCRerr]
IDiscoveryl
Weighted-TermsT0[OCRerr]
5
Pseudo-ThesT0[OCRerr]
D]Mer[OCRerr]e Part-ThesT0[OCRerr]
6 Parsed-Doc Scored-DocT0[OCRerr]
Feature Scoring
t
IPartThesTopI
7 Scored-DocT0[OCRerr] Ranking Top-2000 Scored-Doc(s)T0p
Figure 12: Schematic Pepresentation of Processing When `Pelevant' Documents are Available
advance I (057 1 1.0>
at&t I <057 1 1.0>
bell system breakup I (057 1 2.0>
bell system I <057 1 1.0>
bell I <057 1 1.0>
breakup I <057 1 1.0>
broad statement I <057 1 1.0>
capital spending I <057 1 2.0>
cash [OCRerr]lov I <057 1 2.0>
credit rating I <057 1 2.0>
customer I <057 1 2.0>
cut cost I <057 1 1.0>
cut I <057 1 1.0>
deregulation I <057 1 1.0>
direct indirect result I <057 1 1.0>
Ra bell I <057 1 1.0>
Rci comunication corporation I <057 1 3.0>
mci [OCRerr]inancial health I <057 1 3.0>
mci initiative I <057 1 1.0>
mci I <057 1 2.0>
net income I <057 1 2.0>
net loss I <057 1 2.0>
order I <057 1 1.0>
telecommunication technology I <057 1 1.0>
united states economics I <057 1 1.0>
volume growth I <057 1 2.0>
Figure 13: Sample of Data-i ,201-Term Partitioning Thesaurus for Topic 57
265