LoveHeaven at MAHED 2025: Text-based Hate and Hope Speech Classification Using AraBERT-Twitter Ensemble
Year: 2025
Type: conference-paper
Abstract: This paper presents our system for Sub-task 1 in MAHED 2025 (Zaghouani et al., 2025) shared task: Text-based Hate and Hope Speech Classification.We propose a robust pipeline built upon the bert-base-arabertv02-twitter model, leveraging domain-specific preprocessing, hyperparameter optimization with Optuna, and a K-Fold ensemble strategy.This system ranked 4 th among all participating teams on the leaderboard.We discuss technical design choices, the results of ablation studies, and the impact of preprocessing and model selection on final performance.
Authors: Nguyễn Thiên Bảo, Dang Van Thin
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4416037638