TY - JOUR
T1 - O-Dang at HODI and HaSpeeDe3: A Knowledge-Enhanced Approach to Homotransphobia and Hate Speech Detection in Italian
AU - Di Bonaventura, Chiara
AU - Muti, Arianna
AU - Stranisci, Marco Antonio
N1 - Funding Information:
The work of Chiara Di Bonaventura was supported by UK Research and Innovation [grant number EP/S023356/1], in the UKRI Centre for Doctoral Training in Safe and Trusted Artificial Intelligence (www.safeandtrustedai.org). CDB would like to thank her supervisors, Albert Meroño-Peñuela and Barbara McGillivray, for their helpful comments and mentorship.
Funding Information:
The work of Chiara Di Bonaventura was supported by UK Research and Innovation [grant number EP/S023356/1], in the UKRI Centre for Doctoral Training
Publisher Copyright:
© 2021 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
PY - 2023/9
Y1 - 2023/9
N2 - This paper describes our methods implemented during the EVALITA 2023 campaign for homotransphobia (HODI task) and hate speech detection (HaSpeeDe3 task) in Italian. We present three knowledge-enhanced approaches, namely via triple verbalisation, via prompting and via a majority vote, and we compare them to the AlBERTo baseline. These systems leverage the knowledge graph O-Dang, which contains information about named entities in Italian dangerous speech. Our knowledge-enhanced systems outperformed all the competition's baselines. Our best submissions achieved the macro-F1 score of 0.912 for HaSpeeDe3 and 0.795 for HODI, reaching the 1st and 3rd place, respectively. These results were achieved by using our baseline for HODI, and a majority voting approach for HaSpeeDe3.
AB - This paper describes our methods implemented during the EVALITA 2023 campaign for homotransphobia (HODI task) and hate speech detection (HaSpeeDe3 task) in Italian. We present three knowledge-enhanced approaches, namely via triple verbalisation, via prompting and via a majority vote, and we compare them to the AlBERTo baseline. These systems leverage the knowledge graph O-Dang, which contains information about named entities in Italian dangerous speech. Our knowledge-enhanced systems outperformed all the competition's baselines. Our best submissions achieved the macro-F1 score of 0.912 for HaSpeeDe3 and 0.795 for HODI, reaching the 1st and 3rd place, respectively. These results were achieved by using our baseline for HODI, and a majority voting approach for HaSpeeDe3.
KW - data augmentation
KW - entity linking
KW - hate speech
KW - knowledge graph
KW - prompting
KW - NLP
UR - http://www.scopus.com/inward/record.url?scp=85173573144&partnerID=8YFLogxK
M3 - Article
AN - SCOPUS:85173573144
SN - 1613-0073
VL - 3473
JO - CEUR Workshop Proceedings
JF - CEUR Workshop Proceedings
T2 - 8th Evaluation Campaign of Natural Language Processing and Speech Tools for Italian. Final Workshop, EVALITA 2023
Y2 - 7 September 2023 through 8 September 2023
ER -