User Behavior and Emotion Analysis of Sports Venues Based on Social Media Reviews: The Case of Tianzhu, Shunyi, Beijing

Research on urban space usage based on social media data has attracted growing attention, particularly in urban parks. Sports venues, as important built-environment facilities, have also begun to be examined through online reviews of satisfaction and sentiment, but systematic, theory-driven, multidimensional coding of user behavior and emotion across different venue types remains underexplored. This study extends a park-based analytical framework to eight paid sports venues in Tianzhu, Shunyi, Beijing, using 5317 Dianping and Meituan reviews. A five-dimension analytical framework was constructed covering Physical Activity (PA), Children’s Physical Activity (CPA), Social Interaction (SI), Positive Emotion (PE), and Service Perception (SP), combining manual coding, Tencent word-embedding expansion, and a hybrid matching strategy (jieba segmentation for PA/CPA and substring matching for SI/PE/SP). The main contributions are: (1) a theory-driven multidimensional behavior-emotion coding framework adapted to sports venues; (2) systematic comparison of child-oriented and adult-oriented venue types; (3) operationalization of SP as a first-pass indicator of the institutional/service layer of space production, with its measurement limitations explicitly assessed; and (4) a hybrid matching strategy that accommodates the distinct semantic forms of behavioral and evaluative vocabulary. The study provides transferable analytical tools and evidence-based design and management implications for sports venues as built environments.