Urban parks, as key components of green infrastructure, serve dual functions: ecological services and recreational spaces for residents. With the rapid pace of urbanization and the diversification of public needs, optimizing park planning and management has become an urgent issue. Parks with cultural attributes, such as those along the Grand Canal, not only serve as urban landscapes but also carry significant cultural heritage. The integration of cultural preservation and leisure functions in park planning is essential for maintaining both ecological balance and the cultural identity of a city. Existing research on public perceptions largely focuses on the cultural heritage along the Grand Canal or historical cultural districts, with limited attention given to the role of public perception in the planning of cultural parks along the canal. This paper aims to address this gap by exploring public perceptions of urban parks along the Grand Canal, utilizing network text analysis and machine learning to analyze largescale public commentary data. This study specifically examines parks along the Grand Canal in Suzhou, which is part of the Grand Canal’s cultural heritage and recognized as a UNESCO World Heritage site. These parks are not only recreational areas but also carry deep historical and cultural significance. The research focuses on categorizing and optimizing public perceptions through the analysis of public comments, integrating sentiment analysis, topic modeling, and classification techniques. By capturing the emotional tendencies and experiences of park users, this research aims to inform urban park planning that aligns with public needs and cultural expectations, thereby enhancing the effectiveness of cultural resource utilization and promoting heritage preservation. The data for this study was collected from social media platforms such as Ctrip and Dianping, which provide user-generated content and feedback on parks along the Grand Canal. A total of 29 608 public comments were gathered between 2020 and 2023, focusing on 39 parks. After data cleaning and processing, 15 043 valid comments were retained for analysis. The parks included in the study are categorized into four types: comprehensive parks, specialized parks, community parks, and urban gardens. Using Python-based text mining tools, the comments were processed to extract significant keywords and sentiments. A combination of natural language processing (NLP) techniques, including Jieba word segmentation and TF-IDF (Term Frequency-Inverse Document Frequency) analysis, was employed to extract key perceptual words from the comments. These words were then analyzed using Latent Dirichlet Allocation (LDA) topic modeling to identify prevalentthemes within the comments. The LDA model revealed three primary categories of public perception: 1) public experience and facility management, 2) park design and functional layout, and 3) cultural heritage and community participation. Each category was further subdivided into more specific themes, such as smart facilities, landscape design, and cultural connotations. Sentiment analysis was performed to classify each comment as positive, neutral, or negative based on the presence of sentiment-bearing words. This was followed by an Importance-Performance Analysis (IPA) to identify areas in need of improvement. The analysis revealed several key insights into public perceptions of different park types. Comprehensive and specialized parks were found to face common issues regarding the provision of smart facilities, accessibility for disabled individuals, landscape design, and cultural representation. Community parks, on the other hand, were primarily concerned with service facilities, spatial vitality, and night lighting, while urban gardens faced challenges related to parking, functionality, and a sense of community attachment. The study also employed clustering techniques to identify overlapping issues in parks along the Grand Canal. It was found that comprehensive and specialized parks, due to their larger scale and tourist-focused services, shared concerns related to the integration of smart technologies, accessibility for the disabled, and the adequate display of cultural heritage. Community parks and urban gardens, being smaller in size and closer to residential areas, had distinct issues regarding service facility provision, spatial planning, and the ability to foster a sense of community. Based on these findings, the study provides several recommendations for optimizing urban park planning along the Grand Canal. For comprehensive and specialized parks, improvements in smart technologies, barrier-free facilities, and better integration of cultural heritage in park design are suggested. For community parks, there is a need for enhanced service facilities, such as better lighting, parking, and functional layouts to support community activities. Lastly, urban gardens could benefit from improving their sense of place and offering more practical spaces for local residents. This research underscores the importance of public perception in the planning and development of urban parks. By integrating public feedback through modern data analysis methods, such as network text analysis, this study provides valuable insights for urban planners and policymakers aiming to optimize park spaces for both cultural heritage preservation and community engagement.