This study examines the phenomenon of migration to tsunami-affected areas through the lens of "multiplicity" in digital anthropology, focusing on the case of Minamisanriku Town, Miyagi Prefecture, which was severely impacted by the Great East Japan Earthquake. Employing data from YouTube, Instagram, and Facebook, the research integrates quantitative topic modeling and qualitative ethnographic methods to analyze how migrants express their visions and affects in digital spaces. The findings reveal platform-specific characteristics: YouTube highlights personal experiences and daily life, Instagram emphasizes the sensory appeal and positivity of migration, and Facebook underscores community engagement and collaboration. Using Explainable AI (XAI) tools for image caption analysis, the study uncovers implicit norms and values shaping social behaviors. By comparing digital and face-to-face narratives, this research illuminates how digital platforms mediate self-presentation and social roles, contributing to a nuanced understanding of migration phenomena and advancing methodological innovation in digital anthropology.
