AI-PAVE-Br: Leveraging Large Language Models for Enhanced Product Attribute Value Extraction through a Golden Set Approach

The paper introduces AI-PAVE-Br, a specialized system utilizing Large Language Models (LLMs) for high-accuracy Product Attribute Value Extraction (PAVE) tailored to the Brazilian e-commerce sector. It also presents the Golden Set, a curated dataset with annotated product attributes in Portuguese, which serves as a benchmark for PAVE research. The results demonstrate that AI-PAVE-Br, through targeted prompt engineering, significantly surpasses traditional Named Entity Recognition (NER) methods, providing a scalable solution for non-English markets and contributing valuable resources for NLP research.