User Tools

Site Tools


01_corpus:01_subcorpora

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revisionPrevious revision
Next revision
Previous revision
01_corpus:subcorpora [2019/10/30 16:26] simone01_corpus:01_subcorpora [2022/06/27 09:21] (current) – external edit 127.0.0.1
Line 1: Line 1:
-====== 1.Sub-corpora ====== +====== 1.Sub-corpora ====== 
-Based on the annotation of the languages per chat, the following sub-corpora were created:+Based on the annotation of the languages per chat, different sub-corpora were created.   
 + 
 +The following basic considerations were applied when creating the sub-corpora
 + 
 +===== Definitions for sub-corpora ===== 
 + 
 +  * Each chat was to be assigned to only one language-sub-corpus.  
 +  * Additionally, we differentiate between chats where we have demographic information for all participants and those where we do not. In the former case, the sub-corpus gets the extension _DEMOG. 
 +  * Where additional tasks were performed on individual chats (e.g. normalization or part-of-speech tagging) we created additional sub-corpora per language. 
 + 
 + 
 +===== Main sub-corpora ===== 
 + 
 +  * WUSAll data, i.e. the whole corpus 
 +  * WUS_DEU: All data where non-dialectal German provides the most messages 
 +  * WUS_DEU_DEMOG: A subgroup thereof where we have demographic information from all communication partners. 
 +  * WUS_FRA: All data where French provides the most messages 
 +  * WUS_FRA_DEMOG: A subgroup thereof where we have demographic information from all communication partners. 
 +  * WUS_GSW: All data where dialectal German provides the most messages 
 +  * WUS_GSW_DEMOG: A subgroup thereof where we have demographic information from all communication partners. 
 +  * WUS_ITA: All data where Italian provides the most messages 
 +  * WUS_ITA_DEMOG: A subgroup thereof where we have demographic information from all communication partners. 
 +  * WUS_ROH: All data where Romansh provides the most messages 
 +  * WUS_ROH_DEMOG: A subgroup thereof where we have demographic information from all communication partners. 
 + 
 +Additionally to these corpora, you also see corpora with lowercase letters in the browser (e.g. deu-rftagged, ita-tagged, roh etc.). These corpora contain data from our [[https://wiki.linguistik.uzh.ch/sms4science|SMS project]]. 
 +===== Smaller corpora ===== 
 + 
 +Next to these main sub-corpora, there are some smaller sub-corpora: 
 +  * WUS_SMALL: Chats that are either smaller than 100 messages or where the majority of messages are not in a national language. 
 +  * WUS_SMALL_DEMOG: A subgroup thereof where we have demographic information from all communication partners. 
 +  * WUSdemographics: Only demographic data per person. This sub-corpus is much faster if you want to look up demographic data only. 
 +  * WUS_ARGDROP and WUS_ARGDROP_language: Sub-corpora for which argument drop has been manually annotated. For the architecture of the annotations and scientific considerations behind it see [[http://www.unige.ch/lettres/linge/syntaxe/journal/Volume11/11_Stuntebeck_2018.pdf|Stuntebeck, Franziska (2018): "Annotating Argument Drop in the Swiss WhatsApp Corpus". In: Generative Grammar in Geneva (GG@G) XI, 175-187.]] 
 + 
 + 
 +===== Other corpora in the browsing tool ===== 
 +Additionally to these corpora, you also see corpora with lowercase letters in the browser (e.g. deu-rftagged, ita-tagged, roh etc.). These corpora contain data from our [[https://wiki.linguistik.uzh.ch/sms4science|SMS project]]. 
 + 
 +===== More information about the subcorpora ===== 
 +The individual sub-corpora are well documented in terms of size etc. within the browsing tool. Check the according [[02_browsing:01_sub_corpora|section]] for more information. 
  
01_corpus/01_subcorpora.1572449168.txt.gz · Last modified: 2022/06/27 09:21 (external edit)

Donate Powered by PHP Valid HTML5 Valid CSS Driven by DokuWiki