Eosc/norbert/benchmark

From Nordic Language Processing Laboratory
Revision as of 18:00, 3 December 2020 by Erikve (talk | contribs)
Jump to: navigation, search

Emerging Thoughts on Benchmarking

This would be natural places to start:

NoReC, for document-level sentiment analysis (i.e. rating prediction) https://github.com/ltgoslo/norec

NoReC_fine, for fine-grained sentiment analysis (e.g. predicting target expression + polarity) https://github.com/ltgoslo/norec_fine

NDT, for dependency parsing or PoS tagging (perhaps best to use the UD version) https://www.nb.no/sprakbanken/ressurskatalog/oai-nb-no-sbr-10/

NorNE, for named entity recognition, extends NDT (also available for the UD version) https://github.com/ltgoslo/norne