Two-stage ABSA pipeline in PyTorch & Transformers: extracts product attributes (ATE) and classifies aspect polarities (ASC) using BERT on Women's Clothing and SemEval benchmarks.
python nlp machine-learning ecommerce deep-learning sentiment-analysis text-classification transformers pytorch semeval nlp-machine-learning bert sequence-labeling absa aspect-based-sentiment-analysis text-classification-python huggingface bert-fine-tuning token-classification text-classification-model
-
Updated
Sep 8, 2026 - Jupyter Notebook