Characteristics of Park-related Complaints Identified through Artificial Intelligence in Terms of Ecosystem Services

Purpose: This study aims to analyze park-relatred complaints posted on social media using artificial intelligence (AI)-based natural language processing (NLP) technology to identify the characteristics of these complaints from the perspective of ecosystem services. Method: In this study, Python programming language and its supporting packages were utilized to scrape park-related complaint data registered on social media from July 1, 2023, to June 30, 2024, extracting the words that comprise the complaints. Using AI-based NLP techniques, specifically TF-IDF and K-means clustering algorithms, the importance of the extracted words was calculated and clustered. Additionally, the service support elements of parks inferred from the extracted words were classified into gray infrastructure and green infrastructure, and the characteristics of the complaints were quantified from the perspective of ecosystem services. Result: Complaints related to national parks and urban parks were strongly raised, particularly focusing on park management and infrastructure installation. Using AI to categorize words related to infrastructure into gray and green infrastructure, it was found that most complaints were concentrated on gray infrastructure, such as parking lots and restrooms. Complaints about green infrastructure were relatively few, with some even opposing ecosystem conservation and management. This was interpreted as a result of the public's lack of understanding and interest in the ecosystem services provided by parks.