AI Applications in Emotion Recognition: A Bibliometric Analysis

Abstract

This paper conducts a preliminary exploration of Artificial Intelligence (AI) for emotion recognition, particularly in its business applications. Employing adaptive technologies like machine learning algorithms and computer vision, AI systems analyze human emotions through facial expressions, speech patterns, and physiological signals. Ethical considerations and responsible deployment of these technologies are emphasized through an intense literature review. The study employs a comprehensive bibliometric analysis, utilizing tools such as VOSViewer, to trace the evolution of emotion-aware AI in business. Three key steps involve surveying the literature on emotion analysis, summarizing information on emotion in various contexts, and categorizing methods based on their areas of expertise. Comparative studies on emotion datasets reveal advancements in model fusion methods, exceeding human accuracy and enhancing applications in customer service and market research. The bibliometric analysis sheds light on a shift towards sophisticated, multimodal approaches in emotion recognition research, addressing challenges such as imbalanced datasets and interpretability issues. Visualizations depict keyword distributions in research papers, emphasizing the significance of “emotion recognition” and “deep learning.” The study concludes by offering insights gained from network visualization, showcasing core keywords and their density in research papers. Based on the literature, a SWOT analysis is also conducted to identify the strengths, weaknesses, opportunities, and threats associated with applying emotion recognition to business. Strengths include the technology’s high accuracy and real-time analysis capabilities, enabling diverse applications such as customer service and product quality improvement. However, weaknesses include data bias affecting the AI model’s quality and challenges in processing complex emotional expressions. Opportunities lie in the increasing number of studies, market size, and improving research outcomes, while threats include privacy concerns and growing competition.