Systematic Literature Review on Aspect-Based Sentiment Analysis: Insights from Community Reports

This systematic literature review examines the utilization of datasets, methodologies, and evaluation metrics in Aspect-Based Sentiment Analysis (ABSA) across 44 scholarly publications. Our analysis identifies prevailing trends in dataset selection, with Twitter and TripAdvisor being the most frequently used sources due to their rich and diverse content. We categorize the various algorithms, ranging from lexicon-based approaches like VADER to advanced machine learning and deep learning models such as SVM and BERT. Standard evaluation metrics, including accuracy, F1-Score, precision, and recall, are analyzed to understand how performance is assessed. The review highlights a shift towards deep learning and hybrid models, reflecting recent technological advancements. Our findings offer a comprehensive overview of the current state of ABSA research, identifying existing gaps and providing insights for future studies to enhance the effectiveness of sentiment analysis techniques.