TY - GEN

T1 - A Federated Learning Approach to Privacy Preserving Offensive Language Identification

AU - Zampieri, Marcos

AU - Dola Mullage, Damith

AU - Ranasinghe, Tharindu

N1 - Accepted to TRAC 2024 (Fourth Workshop on Threat, Aggression and Cyberbullying) at LREC-COLING 2024 (The 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation)

PY - 2024/5/20

Y1 - 2024/5/20

N2 - The spread of various forms of offensive speech online is an important concern in social media. While platforms have been investing heavily in ways of coping with this problem, the question of privacy remains largely unaddressed. Models trained to detect offensive language on social media are trained and/or fine-tuned using large amounts of data often stored in centralized servers. Since most social media data originates from end users, we propose a privacy preserving decentralized architecture for identifying offensive language online by introducing Federated Learning (FL) in the context of offensive language identification. FL is a decentralized architecture that allows multiple models to be trained locally without the need for data sharing hence preserving users' privacy. We propose a model fusion approach to perform FL. We trained multiple deep learning models on four publicly available English benchmark datasets (AHSD, HASOC, HateXplain, OLID) and evaluated their performance in detail. We also present initial cross-lingual experiments in English and Spanish. We show that the proposed model fusion approach outperforms baselines in all the datasets while preserving privacy.

AB - The spread of various forms of offensive speech online is an important concern in social media. While platforms have been investing heavily in ways of coping with this problem, the question of privacy remains largely unaddressed. Models trained to detect offensive language on social media are trained and/or fine-tuned using large amounts of data often stored in centralized servers. Since most social media data originates from end users, we propose a privacy preserving decentralized architecture for identifying offensive language online by introducing Federated Learning (FL) in the context of offensive language identification. FL is a decentralized architecture that allows multiple models to be trained locally without the need for data sharing hence preserving users' privacy. We propose a model fusion approach to perform FL. We trained multiple deep learning models on four publicly available English benchmark datasets (AHSD, HASOC, HateXplain, OLID) and evaluated their performance in detail. We also present initial cross-lingual experiments in English and Spanish. We show that the proposed model fusion approach outperforms baselines in all the datasets while preserving privacy.

KW - cs.CL

KW - cs.LG

M3 - Conference contribution/Paper

SN - 9782493814470

BT - TRAC-2024: The Fourth Workshop on Threat, Aggression & Cyberbullying @LREC-COLING-2024

PB - European Language Resources Association (ELRA)

T2 - Fourth Workshop on Threat, Aggression & Cyberbullying @ LREC-COLING-2024

Y2 - 20 May 2024

ER -