A Comprehensive Review of Sentiment Analysis and Aspect-based Sentiment Analysis: Methods, Data Sources, and Open Research Challenges

Authors: Erugu Krishna, Vijay Ramnath Sonawane

Full-text Research PDF File:   View   |   Download

Abstract: Sentiment Analysis (SA) and Aspect-Based Sentiment Analysis (ABSA) have emerged as essential research areas within natural language processing due to the exponential growth of user-generated content on web and social media platforms. These techniques aim to automatically extract, interpret, and classify opinions expressed in textual data, enabling organizations and researchers to understand public perceptions at both coarse and fine-grained levels. This paper presents a comprehensive review of existing sentiment analysis and ABSA approaches, covering traditional lexicon-based methods, classical machine learning algorithms, deep learning models, and recent advances involving transfer learning, contextual embeddings, and multimodal analysis. The study systematically analyzes commonly used data sources, including e-commerce reviews, social media posts, blogs, and multimodal datasets, highlighting their strengths and limitations. Furthermore, the review identifies key challenges such as domain dependency, scarcity of annotated datasets for low-resource languages, handling implicit and context-dependent sentiments, and limited generalization across domains and modalities. By synthesizing findings from recent literature, this paper outlines open research challenges and provides insights into future research directions aimed at developing more robust, scalable, and context-aware sentiment analysis systems.

Keywords: Sentiment Analysis, Aspect-Based Sentiment Analysis, Opinion Mining, Machine Learning, Deep Learning, Multimodal Analysis

Paper Id: 7.711

Published On: 2025-01-15

Published In: Special Issue - International Conference on Engineering, Economics, Management and Applied Sciences (January 2025)

Share this