SENTIMENT DYNAMICS ANALYSIS OF K-POP IDOL AS CORPORATE BRAND AMBASSADOR USING SUPPORT VECTOR MACHINE METHODS AND ISING MODEL

• Main Author: Adha Indraswari, Shabrina
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/73455
• Institution: Institut Teknologi Bandung

Physics is a natural science that aims to explain phenomena through modeling and prediction. This goal is the basis of physics development in a broader field, for the example is complex systems. A complex system is a system consisting of agents who interact with each other in complex ways and can form macroscopic patterns of individual interactions that create new collective behaviors. This phenomenon can be explained in the field of sociophysics. One of the social phenomena that is currently being hotly discussed is the issue of using K-pop idols as brand ambassadors for several companies in Indonesia. Changes in aspects of life that occured due to this phenomenon have caused changes in people's behavior. This is marked by the emergence of various responses regarding the use of K-pop idols as brand ambassadors on social media. The interactions that occur on social media resemble a complex system. These interactions lead to changes or dynamics of the agents’ sentiments in them. In this study, a sentiment analysis of Twitter users' comments regarding the issue of K-pop idols as brand ambassadors was carried out using the Support Vector Machine algorithm. The classification results show that 46% of the dominant data has positive sentiment with the popular keywords "produk", "bangga", "terbesar", and "suka". In addition, it was concluded that the factors of the interactions nearby and extreme event will influence the appearance of changes in sentiment over time (sentiment dynamics). The density factor of positive sentiment and the total population will also affect the time to reach consensus, a condition where sentiment dynamics reach stability