🤖 AI Summary
This study addresses the frequent misclassification of medical terminology and minority-related discourse as harmful content by online moderation systems. Focusing on Bulgarian, the work presents the first toxicity language ontology for the language and introduces a novel fine-grained annotated dataset comprising four categories designed to preserve sensitive yet non-toxic content. By integrating ontology-guided constraints with BERT fine-tuning, the authors train a model on 4,384 manually labeled sentences, achieving a macro-averaged F1 score of 0.89. The proposed approach effectively distinguishes genuinely toxic utterances from critical non-toxic information, offering a deployable solution that significantly enhances both the accuracy and inclusivity of real-world content moderation systems.
📝 Abstract
Toxic content detection in online communication remains a significant challenge, with current solutions often inadvertently blocking valuable information, including medical terms and text related to minority groups. This paper presents a more nu-anced approach to identifying toxicity in Bulgarian text while preserving access to essential information. The research explores two distinct methodologies for detecting toxic content. The developed methodologies have po-tential applications across diverse online platforms and content moderation systems. First, we propose an ontology that models the potentially toxic words in Bulgarian language. Then, we compose a dataset that comprises 4,384 manually anno-tated sentences from Bulgarian online forums across four categories: toxic language, medical terminology, non-toxic lan-guage, and terms related to minority communities. We then train a BERT-based model for toxic language classification, which reaches a 0.89 F1 macro score. The trained model is directly applicable in a real environment and can be integrated as a com-ponent of toxic content detection systems.