BERT-Based Joint Model for Aspect Term Extraction and Aspect Polarity Detection in Arabic Text

Aspect-based sentiment analysis (ABSA) is a method used to identify the aspects discussed in given text and determine expressed towards each aspect. This can help provide more fine-grained understanding of opinions text. The majority Arabic ABSA techniques use today significantly rely on repeated pre-processing feature-engineering operations, as well outside resources (e.g., lexicons). In essence, there significant research gap NLP with regard transfer learning (TL) language models for aspect term extraction (ATE) polarity detection (APD) While TL has proven be an effective approach variety tasks other languages, its context been relatively under-explored. paper aims address this by presenting TL-based ATE APD Arabic, leveraging knowledge capabilities previously trained models. base (Arabic version) BERT model serves foundation suggested Different implementations are also contrasted. A reference dataset was experiments (HAAD dataset). experimental results demonstrate that our surpass baseline proposed approaches.