Identifying Offensive Expressions of Opinion in Context
This addresses the lack of resources for extracting subjective, offensive information in NLP, benefiting researchers and practitioners in sentiment analysis and hate speech detection, though it is incremental as it builds on existing lexicon and annotation methods.
The paper tackles the challenge of identifying offensive opinions in context by creating a cross-lingual offensive lexicon with explicit and implicit expressions, annotated for context dependency and hate speech markers, achieving high human inter-annotator agreement.
Classic information extraction techniques consist in building questions and answers about the facts. Indeed, it is still a challenge to subjective information extraction systems to identify opinions and feelings in context. In sentiment-based NLP tasks, there are few resources to information extraction, above all offensive or hateful opinions in context. To fill this important gap, this short paper provides a new cross-lingual and contextual offensive lexicon, which consists of explicit and implicit offensive and swearing expressions of opinion, which were annotated in two different classes: context dependent and context-independent offensive. In addition, we provide markers to identify hate speech. Annotation approach was evaluated at the expression-level and achieves high human inter-annotator agreement. The provided offensive lexicon is available in Portuguese and English languages.