COMMUNITY STRUCTURE DETECTION OF INFLUENCERS ON YOUTUBE PLATFORM USING COMBINED METHODS OF GENETICS ALGORITHM AND SIMULATED ANNEALING

• Main Author: Dinar Latiffa, Ilma
• Format: Final Project
• Language: Indonesia
• Subjects: Fisika
• Online Access: https://digilib.itb.ac.id/gdl/view/41879
• Institution: Institut Teknologi Bandung

The development of more physics give birth to complex system. One of the topic of complex system that has been developed recently is sociophysics. Community detection has been widely accepted as one of sociophysics developing research. One of the most important in complex system is the interaction of each components. This interaction can be represented by using a graph or network. This graph or network consist from many nodes and the connection of each nodes called edges. If the individual are the nodes then community is a group that consists of many individual that can interact with each other but hardly interact with other individual from other community. Community detection has many benefits to comprehend about connection and structure on human society. Genetic algorithm is the method that based on genetics process on living being. To enhance the results from genetic algorithm, Simulated Annealing method will be used too. To know how bad or how good the result is, the concept of modularity will be used. In this final project, the genetic algorithm and Simulated Annealing will be used as method to detecting community. The community in this final project is the top 100 most subscribed Indonesian youtuber and has collaborated with minimal one youtuber. In this final project there are 4 community.