Public perceptions and expectations of urban park environments on children's play: A GIS and text mining analysis on children's activity in Atlanta's parks using social media data

Outdoor activities confer mental, physical, and social health benefits to children. In urban areas, parks are vital for children to engage in outdoor activities. Understanding public perceptions and expectations of how urban park environments influence children's activity could attenuate social and physical inequities and promote park use. Previous studies have used surveys and observational data to analyze preferences for children's activity in parks. Crowdsourced online data and volunteered geographic information (VGI) could offer a valuable addition to or substitute for traditional approaches. Our study uses social media as a passive data collection method to ascertain public perceptions and expectations about children's activity in Atlanta parks with a shorter timeframe, compared to a regular survey analysis. We collected 4026 Google Map reviews from 2017 to 2022 that related to children and applied text mining analysis to understand how people perceive children's activity in Atlanta parks and how perceptions change across different environments. Ten topics were extracted that related to children's activity in Atlanta parks: attitude, amenity, safety, social, pet, nature, recreational, water sports, water recreational, and sports. The attitude, social, and water recreational topics were the most widely discussed topics in all settings. However, in racially diverse and low-crime environments, discussions centered on pet, recreational, and sports, suggest the importance of tailored strategies to promote children's activities in parks. Park planners and policymakers can use this approach and findings to evaluate and advocate for children's activities in urban parks. Also, our work helps to expand survey analysis with passive data collection methods from small geographic scales to larger areas and to apply geoanalytics of big data and social media data in investigation research.