TUdatalib : DBS Corpus

Zur Kurzanzeige

dc.contributor.author	Benikova, Darina
dc.contributor.author	Mieskes, Margot
dc.contributor.author	Meyer, Christian M.
dc.contributor.author	Gurevych, Iryna
dc.date.accessioned	2019-02-18T12:19:55Z
dc.date.available	2019-02-18T09:28:24Z
dc.date.available	2019-02-18T12:19:55Z
dc.date.issued	2017-05-29
dc.identifier.uri	https://tudatalib.ulb.tu-darmstadt.de/handle/tudatalib/1915.2
dc.description	The DBS corpus contains 93 multi-document summaries for 293 German documents about 30 education-related topics. We sampled the topics from the Deutscher Bildungsserver (DBS) webpage and crawled the documents linked there. The documents are highly heterogeneous in terms of text type, genre, and style. The multi-document summaries are the result of a seven step annotation process yielding coherent extracts – a novel type of summary that is based on phrases extracted from the original documents that have been ordered and minimally redacted to form a well-readable, coherent text. The data of all intermediate steps is part of the repository to allow for extensive system evaluation. If you use the corpus in academic works, please cite our COLING paper.	en_US
dc.language.iso	de	en_US
dc.relation	IsDescribedBy;URL;https://aclweb.org/anthology/C16-1099
dc.relation	References;URL;https://www.bildungsserver.de/
dc.relation	IsVersionOf;URL;https://github.com/AIPHES/DBS
dc.rights	Creative Commons Attribution Share-Alike 4.0
dc.rights.uri	https://creativecommons.org/licenses/by-sa/4.0/
dc.subject	Multi-document Summarization	en_US
dc.subject	Heterogeneous Sources	en_US
dc.subject	Information Aggregation	en_US
dc.subject	Natural Language Processing	en_US
dc.subject	AIPHES	en_US
dc.subject.ddc	000 Informatik, Informationswissenschaft, allgemeine Werke	en_US
dc.subject.ddc	430 Germanische Sprachen; Deutsch	en_US
dc.title	DBS Corpus	en_US
dc.type	Dataset	en_US
dc.type	Text	en_US
dc.type	Workflow	en_US
dc.description.version	2.0
tud.tubiblio	97945