ANALYSIS OF SENTIMENT DYNAMICS REGARDING MARKETPLACES DUE TO THE CLOSURE OF TIKTOK SHOP USING THE SUPPORT VECTOR MACHINE METHODS AND ISING MODELS.

• Main Author: Yusuf Firdaus, Ahmad
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/81249
• Institution: Institut Teknologi Bandung

Physics is a fundamental science in studying the behavior of the universe, enabling the exploration of various aspects of life, such as sociophysics. Sociophysics is a branch of physics used as an analogy or methodological tool to analyze the fundamental laws and principles that explain human behavior, thereby determining collective social phenomena arising from interactions between individuals as the basic components of social structure. One of the current social phenomena is the dynamics of Indonesian public sentiment related to marketplaces following the closure of TikTok Shop. The Indonesian government's closure of TikTok Shop in October sparked debates among Indonesians on social media, resulting in a dynamic sentiment phenomenon. I conducted a study to analyze the sentiment dynamics through sentiment analysis of social media X / Twitter users using Machine Learning Support Vector Machine and Ising Model Simulation. My research results indicate that Indonesian public sentiment regarding marketplaces from September to December was predominantly negative, with a rate of 45.8%. The distribution of sentiments changed daily, reflecting the dynamic sentiment phenomenon. This phenomenon was influenced by several factors: the initial sentiment density, the population involved, and the occurrence rate of triggering events. These factors affect the average sentiment value within a system and the time it takes for the system to reach equilibrium or stability.