(090) previous ~ index ~ next
To: tdt-distrib@unagi.cis.upenn.edu
From: James Allan <allan@cs.umass.edu>
Subject: Topic relevance error
Date: Wed, 15 Jul 1998 17:35:07 -0400
TDTers,
I suppose this is mostly for the folks at LDC. Ron Papka here at
UMass has been poking around with the TDT2 corpus and stumbled upon
what looks like an oddity in the data. This is from the 5/22 version
of the Jan-Feb corpus. He noticed four out of whack judgements:
topic docid judgement
2 TDT00300 APW19980105.0021 YES
2 TDT00332 APW19980105.0549 YES
2 TDT00333 APW19980105.0550 YES
2 TDT00375 APW19980105.0808 YES
are about the decline of the Baht in Thailand, and topic
2 is about Monica Lewinsky. Unfortunately these docs are in
the judgments file under topic 2.
(The TDTnnnnn are internal for our use.) I don't see any other topic
that has anything to do with Thailand, so I'm not sure what's up here.
On the assumption that these are relevance judgement errors that need
to be corrected, I'm passing them along.
-- james
(090) previous ~ index ~ next
Last updated Wed Sep 9 09:40:52 1998