Prior to the creation of this publication, NIST had been distributing the text
collection via ftp to TREC participants but the size of the collection made ftp
transfer cumbersome.  The LDC has moved the TREC Mandarin text collection to
CD-ROM to simplify its distribution.

This single CD publication contains an index.html file with links to the
peoples-daily/ and xinhua/ directories containing newswire text and to a doc/
directory containing additional documents such as this file, file.tbl.  

The 36 files in the peoples-daily directory contain 139,801 documents; the 41
files in the xinhua directory contain 24,988 documents.  The contents of the
peoples-daily and xinhua directories are shown below.  The files in the
poeples-daily directory are named as pdYYMM.sgml where pd = People's Daily, YY
= year, and MM = month.  The files in the xinhua directory are named as
xYYMMddd where x = Xinhua, YY = year, MM = month, and ddd = sequential order
number.

/index.html

doc/

file.tbl                    
pd.dtd
pd-char-err.log             
pd-char-err.summary         
pd-missed-boundary.summary  
xh-char-err.log
xh-char-err.summary
xinhua.dtd

peoples-daily/

pd9101.sgml
pd9102.sgml
pd9103.sgml
pd9104.sgml
pd9105.sgml
pd9106.sgml
pd9107.sgml
pd9108.sgml
pd9109.sgml
pd9110.sgml
pd9111.sgml
pd9112.sgml
pd9201.sgml
pd9202.sgml
pd9203.sgml
pd9204.sgml
pd9205.sgml
pd9206.sgml
pd9207.sgml
pd9208.sgml
pd9209.sgml
pd9210.sgml
pd9211.sgml
pd9212.sgml
pd9301.sgml
pd9302.sgml
pd9303.sgml
pd9304.sgml
pd9305.sgml
pd9306.sgml
pd9307.sgml
pd9308.sgml
pd9309.sgml
pd9310.sgml
pd9311.sgml
pd9312.sgml

xinhua/

x9404001
x9405001
x9406001
x9406002
x9406003
x9407001
x9407002
x9407003
x9408001
x9408002
x9408003
x9409001
x9409002
x9409003
x9410001
x9410002
x9410003
x9501001
x9501002
x9501003
x9502001
x9502002
x9502003
x9503001
x9503002
x9503003
x9504001
x9504002
x9504003
x9505001
x9505002
x9505003
x9507001
x9507002
x9507003
x9508001
x9508002
x9508003
x9509001
x9509002
x9509003