This is a description of the corpora held by the
Arbeitsbereich Linguistik, University of M"unster. The data are in
accordance  with the survey questionaire sent out  by the Center for
Electronic Texts in the Humanities.
If you need further information, please contact:

Prof. Dr. Wolf Paprott'e
University of Muenster - Arbeitsbereich Linguistik
Huefferstr. 27
W-4400 MUenster
e-mail: paprott@dmswwu1a.bitnet

or, for technical details,

Lothar Lemnitzer
(same address)
e-mail: lothar@morrison.uni-muenster.de


DATA STRUCTURE


language <la>
source   <so>
text type	<tt>
size in running words <rw>
size in MB	<mb>
representation format  <re>
availability    <av>
conversion tools used	<cv>
current uses	<cu>

THE GERMAN CORPUS


<la> german
<so> FAZ
<tt> newspaper texts
<rw> ca. 80 mill.
<mb> ca. 210 
<re> SGML, DTD
<av> In general, NO. Access for specific tasks possible
<cv> conversion from typesetting tape to SGML representation. Implemented  
     in C and lex.
<cu> statistical analyses, lexicographic tasks (lemma selection, extraction
     of example sentences)

           
<la> german
<so> ZEIT
<tt> newspaper texts 
<rw> ca. 14 mill
<mb> ca. 90 
<re> SGML, according to own DTD
<av> In general, NO. Access for specific tasks possible
<cv> conversion from typesetting tape to SGML representation. Implemented  
     in C and lex.
<cu> statistical analyses, lexicographic tasks (lemma selection, extraction
    of example sentences)


<la> german
<so> VDI-Nachrichten
<tt> newspaper texts
<rw> 192 000
<mb> 1.5 
<re> SGML, according to own DTD
<av> YES
<cv> conversion from typesetting tape to SGML representation. Implemented  
     in C and lex. Incomplete
<cu> statistical analyses, lexicographic tasks (lemma selection, extraction
   of example sentences)


<la> german
<so> die Bibel - Elberfelder Ausgabe 
<tt> prose 
<rw> 780 000 
<mb> 4.5 
<re> plain ASCII files 
<av> YES
<cv> none

<la> german
<so> Siemens 
<tt> technical documentations 
<rw> 100 000 
<mb> 0.5 
<re> plain ASCII files 
<av> YES 
<cv> none 
<cu> statistical analyses, lexicographic tasks (lemma selection, extraction
   of example sentences)

Furthermore, we collect scientific works and papers. Up to now, we hold:

1 dissertation in medicine
2 M.A. theses
some linguistic articles


THE ENGLISH CORPUS

<la> english
<so> Oxford Text Archive
<tt> Brown Corpus, misc.
<rw> 1. mill
<mb> ?
<re> plain ASCII files
<av> YES
<cv> none
<cu> none

<la> english
<so> Longman
<tt> Longman Lancaster corpus, misc.
<rw> ca. 10 mill.
<mb> ?
<re> plain ASCII files
<av> YES
<cv> none
<cu> none


<la> english
<so> St. James Bible 
<tt> prose 
<rw> 832 000 
<mb> 4.5 
<re> plain ASCII files 
<av> YES 
<cv> none 
<cu> none

<la> english
<so> Dylan texts 
<tt> prose, interviews 
<rw> 10 000 
<mb> 0.06 
<re> plain ASCII files 
<av> YES 
<cv> none 
<cu> none

<la> english
<so> Hansard corpus (English part) 
<tt> discussion protocols, legal/political texts 
<rw> ? 
<mb> ? 
<re> plain ASCII files 
<av> YES. 
<cv> none 
<cu> none

<la> english
<so> Usenet - discussion about Japanese politics 
<tt>  scientific prose  
<rw> 200 000 
<mb> 1.3 
<re> plain ASCII files 
<av> YES 
<cv> none 
<cu> none


<la> english
<so> Usenet - discussions  - politics and recreation
<tt>  prose  
<rw> 180 000 
<mb> 1.2 
<re> plain ASCII files 
<av> YES 
<cv> none 
<cu> none


<la> english
<so> Usenet - science fiction discussion 
<tt>  prose  
<rw> 4.15 mill 
<mb> 26 
<re> plain ASCII files 
<av> YES 
<cv> none 
<cu> none


<la> english
<so> Usenet - discussions  - telecommunications 
<tt> prose  
<rw> 3.5 mill 
<mb> 22.5 
<re> plain ASCII files 
<av> YES 
<cv> none 
<cu> none


<la> english
<so> Usenet - discussions  - telecommunications 
<tt> prose  
<rw> 3.5 mill 
<mb> 22.5 
<re> plain ASCII files 
<av> YES 
<cv> none 
<cu> none

<la> english
<so> ACL/DCI Cd-ROM 
<tt> Newspaper texts, abstracts, dictionary  
<rw> 65 mill 
<mb> ? 
<re> almost all texts in SGML 
<av> YES 
<cv> none 
<cu> statistical analyses, lexicographic tasks (lemma selection, extraction
   of example sentences)


SPANISH CORPUS


<la> spanish 
<so> el diario vasco 
<tt> newspaper texts
<rw> 830 000 
<mb> 6.5 
<re> plain ASCII files 
<av> YES 
<cv> none  
<cu> statistical analyses, lexicographic tasks (lemma selection, extraction
   of example sentences)

<la> spanish 
<so> el pais 
<tt> newspaper texts
<rw> ca. 24 mill 
<mb> ca. 140 
<re> SGML-like 
<av> NO 
<cv> conversion from typesetting tape to SGML representation. Implemented
     in C and lex. Incomplete
<cu> statistical analyses, lexicographic tasks (lemma selection, extraction
   of example sentences)

<la> spanish 
<so> abc 
<tt> newspaper texts
<rw> ?  
<mb> 3 
<re> plain ASCII files 
<av> NO 
<cv> none  
<cu> none

<la> spanish 
<so> private 
<tt> scientific prose 
<rw> 40 000  
<mb> 0.22  
<re> plain ASCII files 
<av> YES 
<cv> none  
<cu> statistical analyses, lexicographic tasks (lemma selection, extraction
   of example sentences)

<la> spanish 
<so> private 
<tt> fiction (contemporary) 
<rw> 750 000 
<mb> 5 
<re> plain ASCII files 
<av> YES 
<cv> none  
<cu> statistical analyses, lexicographic tasks (lemma selection, extraction
   of example sentences)


FRENCH CORPUS


<la> french 
<so> ECI 
<tt> hansard corpus (French part) 
<rw> ca 20 mill
<mb> ? 
<re> plain ASCII files 
<av> YES.  
<cv> none  
<cu> statistical analyses, lexicographic tasks (lemma selection, extraction
   of example sentences)




ITALIAN CORPUS


<la> italian 
<so> Univ. of Pisa 
<tt> newspaper texts, prose 
<rw> 370 000 
<mb> 1.4 
<re> plain ASCII files 
<av> YES. 
<cv> none  
<cu> lexicographic tasks (lemma selection)


Material of two Italian newspapers (il messagiero, 24ore) is under work
now, so not yet available