Aspect-Based Approaches for Measuring Customer Feedback in the E-Commerce Industry

This study provides an overview of how sentiment analysis has evolved in the e-commerce industry, focusing specifically on aspect-based approaches. These approaches offer a more detailed way of analyzing customer opinions by considering specific aspects of products or services. The study explores various aspect-based sentiment analysis (ABSA) methods used in e-commerce, highlighting both their advantages and limitations. ABSA models are categorized into three main types: knowledge-based, machine-learning, and hybrid models. The study also examines the latest ABSA models, including those that combine deep learning techniques with knowledge bases and transformer models for improved analysis.