NorGramBank is a parsebank of Norwegian that is under development in the INESS project. It covers the two written standards of Norwegian (Bokmål and Nynorsk) and has varied types of texts, both fiction (adult and children’s fiction) and non-fiction (newspapers, information brochures, research articles, etc).
NorGramBank is being built by automatically parsing with NorGram, a hand-written broad coverage computational grammar. NorGram is written in the Lexical Functional Grammar (LFG) framework on the XLE (Xerox Linguistic Environment) platform. It provides detailed grammatical analyses on two levels, constituent structure and functional structure. Parsing is done with XLE, and the LFG Parsebanker is used for semi-automatic disambiguation (Rosén et al. 2012, 2009, etc., see the Publications page at http://clarino.uib.no/iness).
Parts of the material are annotated manually; the rest is disambiguated using statistical parse ranking. This is work in progress, and the portion of manually annotated material may increase.
The full source texts remain copyright protected, and cannot be redistributed by INESS. The full list of authors and texts in each treebank can be found in INESS via "Treebank overview" in the left-hand menu in the INESS portal.
The link will take you to an external site: We take no responsibility whatsoever for the content of external links.