Social Network Analysis and Sentiment Classification of Extended Reality Product Content

Yerik Afrianto Singgalen (Universitas Katolik Indonesia Atma Jaya, Jakarta)

KLIK: Kajian Ilmiah Informatika dan Komputer
Vol. 4 No. 4 (2024): Februari 2024

This study explores Extended Reality (XR) products, specifically focusing on the Apple Vision Pro, to elucidate consumer perceptions and the underlying social dynamics of these innovative technologies. This research delves into Extended Reality (XR) products, specifically focusing on the Apple Vision Pro, aiming to understand consumer perceptions and social dynamics surrounding these innovative technologies. By leveraging sentiment analysis and Social Network Analysis (SNA) alongside CRISP-DM and SVM algorithms, this study provides a comprehensive insight into sentiment patterns, network structures, and influential factors within the XR community. A multi-faceted approach is adopted to achieve the research objectives. Sentiment analysis and SNA dissect sentiment patterns and uncover network structures within the XR community. The CRISP-DM framework guides the research process, ensuring systematic data analysis and interpretation. SVM algorithms classify sentiments, providing a robust analytical framework for understanding consumer sentiments towards XR products. The analysis yields significant insights into XR consumer perceptions and social dynamics. The calculated network metrics, including a density of 0.000124, absence of reciprocity, centralization value of 0.001331, and modularity value of 0.999000, shed light on crucial network dynamics within the XR community. Examining frequently used words reveals prevalent topics within the XR discourse, providing valuable context for understanding consumer sentiments. Furthermore, the evaluation of SVM algorithms demonstrates commendable performance metrics, with the SVM without SMOTE achieving an accuracy rate of 84.33%, precision of 84.67%, recall of 99.28%, and f_measure of 91.39%. In comparison, the SVM with SMOTE exhibits an accuracy of 81.82% and a precision of 97.58%. This research contributes valuable insights into the consumer landscape of XR products, mainly focusing on the Apple Vision Pro. By combining sentiment analysis, SNA, and established methodologies, the study offers a nuanced understanding of consumer perceptions and social dynamics within the XR community. These findings inform strategic decisions and contribute to advancements in XR technologies, offering valuable insights into the efficacy of sentiment analysis techniques in understanding consumer sentiments