(109) previous ~ index ~ next
To: tdt-distrib@unagi.cis.upenn.edu
From: Sreenivasa Sista <ssista@bbn.com>
Subject: TDT2 July release - ASR TEXT - doc has no data - judged as "yes"
Date: Tue, 4 Aug 1998 14:48:16 -0400 (EDT)
In the TDT2 July release, ASR Text, docno CNN19980410.1130.1333 (topic =
66), we do not have any asr data for the defined boundaries. Whereas, the
document is judged as "yes" (which is true if we look at the corresponding
news wire sgml file). This is a training story i.e., it is as good as
having (Nt-1) training documents.
------------------------------------------
contents of 19980410_1130_1200_CNN_HDL.bndasr
****<BOUNDARY docno=CNN19980410.1130.1333 doctype=NEWS Bsec=1333.98
Esec=1344.60>***********
<BOUNDARY docno=CNN19980410.1130.1344 doctype=NEWS Bsec=1344.60
Esec=1353.62 Brecid=3375 Erecid=3406>
-----------------------------
contents of topic_relevance.table
<ONTOPIC topicid=66 level=YES docno=CNN19980410.1130.1333
fileid=19980410_1130_1200_CNN_HDL comments=NO>
-----------------------------
contents of 19980410_1130_1200_CNN_HDL.asr
<W recid=3373 Bsec=1321.77 Dur=0.32 Clust=30 Conf=0.77> FIVE
*********<X Bsec=1322.09 Dur=22.59 Conf=NA>************
<W recid=3374 Bsec=1344.68 Dur=0.11 Clust=23 Conf=0.82> TO
-----------------------------
What are we supposed to do in these conditions ?
Sreenivasa P. Sista
ssista@bbn.com
(109) previous ~ index ~ next
Last updated Wed Sep 9 09:40:53 1998