cardiffnlp/twitter-xlm-roberta-base-hate-spanish

This model is a fine-tuned version of cardiffnlp/twitter-xlm-roberta-base using the HaterNet dataset and the Spanish subset of SemEval-2019 Task 5.

Following metrics are achieved

• on the test split of SemEval-2019 Task 5

  • F1 (weighted): 0.7866
  • F1 (macro): 0.7935
  • Accuracy: 0.7937

• on custom test split of Haternet

  • F1 (weighted): 0.7815
  • F1 (macro): 0.6981
  • Accuracy: 0.7933

• on Haternet & SemEval-2019 Task 5

  • F1 (weighted): 0.7908
  • F1 (macro): 0.7657
  • Accuracy: 0.7936

Usage

Install tweetnlp via pip.

pip install tweetnlp

Load the model in python.

import tweetnlp
model = tweetnlp.Classifier("cardiffnlp/twitter-xlm-roberta-base-hate-spanish")
model.predict('Ismael es egocentrico porque se vuelve loca si le dicen que tiene el pelo bonito😂😂😂😂 eso se define con otro objetivo #FirstDates251')
>> {'label': 'NOT-HATE'}

Datasets

@inproceedings{basile-etal-2019-semeval, title = "{S}em{E}val-2019 Task 5: Multilingual Detection of Hate Speech Against Immigrants and Women in {T}witter", author = "Basile, Valerio and Bosco, Cristina and Fersini, Elisabetta and Nozza, Debora and Patti, Viviana and Rangel Pardo, Francisco Manuel and Rosso, Paolo and Sanguinetti, Manuela", booktitle = "Proceedings of the 13th International Workshop on Semantic Evaluation", month = jun, year = "2019", address = "Minneapolis, Minnesota, USA", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/S19-2007", doi = "10.18653/v1/S19-2007", pages = "54--63", }

@article{quijano2019haternet, title={HaterNet a system for detecting and analyzing hate speech in Twitter (Version 1.0)[Data set]}, author={Quijano-Sanchez, Lara and Kohatsu, Juan Carlos Pereira and Liberatore, Federico and Camacho-Collados, Miguel}, journal={Zenodo}, year={2019} }