This study applies Latent Dirichlet Allocation (LDA) and Word2Vec-based semantic network analysis to explore the elements that contribute to the perception of Animal Crossing: New Horizons as a healing game. User comments were collected, with LDA used to extract latent topics and Word2Vec to analyze semantic associations among key terms. The analysis identified major themes such as gameplay immersion, resource crafting and settlement activities, emotional satisfaction, restoration of daily rhythms, and formation of social relationships. Word2Vec further revealed detailed emotional experiences and activity contexts around key terms like island, craft, relaxing, time, and people. The integrated findings confirmed that Animal Crossing fosters psychological stability and healing experiences across both structural (e.g., spatial creation, task routines) and emotional (e.g., relaxation, social bonding) dimensions. This study also highlights how LDA and Word2Vec can be complementarily applied to user-generated text analysis, demonstrating the methodological potential for expanding natural language processing-based approaches in game user experience research.

ABSTRACT
1. Introduction
2. Background
    2.1 Introduction of Animal Crossing: New Horizons
    2.2 LDA
    2.3 Word2Vec
3. Experimental Method
    3.1 Collecting Data
    3.2 Data Preprocessing
    3.3 Applying a LDA model
    3.4 Applying a Word2Vec model
4. Review Analysis Result
    4.1 Key topics with LDA
    4.2 Word2Vec-based Semantic Network Analysis
5. LDA and Word2Vec Integrated Analysis
6. Conclusion
참고문헌

• Beomjun JO(Department of Game Design and Development, Graduate School, Sangmyung University, Seoul, South Korea) | 조범준
• Taesuk KIHL(Department of Game Design and Development, Graduate School, Sangmyung University, Seoul, South Korea) | 길태숙 Corresponding author