PyTorch model for Slovenian Coreference Resolution

  Authors

Klemen, Matej ; Čebular, Martin and Žitnik, Slavko

  Item identifier

http://hdl.handle.net/11356/1773

 Project URL

https://rsdo.slovenscina.eu/en/semantic-resources-and-technologies

 Demo URL

https://slovenscina.eu/odkrivanje-koreferencnosti

 Referenced by

https://doi.org/10.2298/CSIS201120060K

 Date issued

2023-02-17

 Type

toolService

 Language(s)

Slovenian

 Description

Slovenian model for coreference resolution: a neural network based on a customized transformer architecture, usable with the code published on https://github.com/matejklemen/slovene-coreference-resolution. The model is based on the Slovenian CroSloEngual BERT 1.1 model (http://hdl.handle.net/11356/1330). It was trained on the SUK 1.0 training corpus (http://hdl.handle.net/11356/1747), specifically the SentiCoref subcorpus. Using the evaluation setting where entity mentions are assumed to be correctly pre-detected, the model achieves the following metric values: MUC: precision = 0.931, recall = 0.957, F1 = 0.943 BCubed: precision = 0.887, recall = 0.947, F1 = 0.914 CEAFe: precision = 0.945, recall = 0.893, F1 = 0.916 CoNLL-12: precision = 0.921, recall = 0.932, F1 = 0.924

 Publisher

Faculty of Computer and Information Science, University of Ljubljana

 Acknowledgement

Ministry of Culture C3340-20-278001 "Development of Slovene in a Digital Environment"

 Subject(s)

Slovenian coreference resolution neural networks

 Collection(s)

CLARIN.SI data & tools

Show full item record