ABSTRACT
Parks play a critical role in the health and well-being of people in urban environments. Recent events such as the COVID-19 pandemic have underscored the importance of public parks, particularly in cities. To maximize the benefits of these spaces, it is important to understand the social dimension of site user experiences within successful urban parks. Social media data provide a means to assess public places through the lens of large quantities of site users over time. Recent landscape research studies provide assessments of urban and nature-based locations using social media data and are predominantly quantitative in their methodologies (Havinga et al., 2021;Donahue et al., 2018;Wood et al., 2013). However, incorporating mixed methods into existing approaches can create a better understanding of site user experiences. This study uses 11,419 Tripadvisor reviews from the years 2010 to 2018 in a multi-step process, where qualitative content analysis builds upon Latent Dirichlet allocation (LDA) modeling topics to assess the semantic content of Bryant Park, NY. The new framework emerging from this process separates site user perceptions and sentiment positivity into categories. Site design elements are revealed as a major positive focus of site users, along with the position of the park within the urban fabric, site activity consisting of passive pastimes, and the essence of the space related to emotions. The study's findings can provide guidance for designers and park managers for the creation of successful urban parks and provide another baseline for research on New York City's parks.