Anti-Korean Sentiment and Collective Meaning-Making in Platform Comments : An Analysis of Japanese-Language YouTube Comments on the Hosei University “Hammer Woman” Incident

Key Points

• To investigate how anti-Korean sentiment and collective meaning-making formed across Japanese-language YouTube comments following a university hammer assault incident.
• Analyzed 6,272 Japanese-language comments across 131 YouTube videos regarding the January 2025 Hosei University assault incident.
• Conducted GiNZA-based Japanese morphological analysis, TF-IDF weighting, co-occurrence semantic network analysis with Louvain community detection, and Latent Dirichlet Allocation (LDA) topic modeling.
• Semantic network analysis identified four major vocabulary communities, highlighting a core cluster that directly linked negative representations with national and ethnic categorizations.
• LDA topic modeling extracted six latent topics encompassing perpetrator motive attribution, intersectional nationality and gender categorizations, and the politicization of foreigner exclusion.
• User discourse frequently escalated from individual perpetrator condemnation into generalized anti-Korean prejudice, broader xenophobic claims, and government policy critiques.