NoSta-D -- Korpus von Nicht-Standardvarietäten des Deutschen

Corpus of different varieties of German. The subcorpora are subsets of other corpora, specified in parentheses: 1.) historical data (Anselm Corpus), chat data (Dortmund Chat Corpus), learner data (Falko), spoken data (BeMaTaC), literary prose (Kafka); 2.) newspaper texts (TüBa-D/Z). The subcorpora chat, spoken data, prose, and newspaper consist of approximately 5,000 tokens each, historical data of 1,000 tokens, and learner data of 2,900 tokens. Each subcorpus is annotated with the following information: token and sentence boundaries; normalization; POS tags and dependency relations; named entities; coreference.