Objective With the rapid advancement of urbanization and the rising expectations of citizens for high-quality living environments, the Beijing Municipal People’s Government officially promulgated the Special Plan for Beijing Garden City (2023−2035) in May 2024, in response to these emerging needs. This comprehensive urban planning initiative outlines the strategic objective of achieving “overall colorization” centered around the development and enhancement of urban parks. The policy emphasizes the central role of flowers in landscape construction, ecological restoration, and promoting livable, beautiful, and culturally vibrant urban habitats. Beyond their ecological and visual value, flowers also serve as cultural symbols and contribute to residents’ emotional well-being. However, the effectiveness of floral landscape construction depends heavily on public perception, which influences the usage of these green spaces. This research investigates public perceptions of floral landscapes in urban parks across the central urban area of Beijing, with a view to identifying key themes, species, and perception patterns across different park types, thereby offering practical insights for improving landscape planning and supporting the broader goals of garden city development.

Method Taking 78 urban parks in the central urban area of Beijing as the research objects, this research uses Python to collect annual reviews for each park from dianping.com in 2023, yielding 13,657 valid entries, and a combination of content analysis (CA), latent Dirichlet allocation (LDA) topic modeling and importance − performance analysis (IPA) is employed to examine and evaluate public perceptions of floral landscapes in urban parks across Beijing. The specific process involves the following steps: Constructing a customized lexicon and segmenting the review data; establishing a flower plant perception corpus and performing CA and word frequency analysis to assess the overall perception of flower plants; applying the LDA model to extract thematic clusters and corresponding keywords, and summarizing the perception themes in alignment with the Garden City planning framework; evaluating the importance and satisfaction levels of the identified perception themes using the IPA method; and finally, categorizing and discussing the findings to propose strategies for enhancing public perception of floral landscapes.

Result A list of the top 15 flower plant species based on public perception is compiled, and four flower perception themes are identified. Significant differences are found in public perceptions across different park types. 1) In terms of overall perception, there are considerable differences in the degree of public perception and spatial distribution among various types of parks. Time patterns are clearly defined, with perception levels peaking primarily between March and May and September and November. A wide variety of flower species are employed, with lotus (2,250 mentions), peach blossom (1,376 mentions), and cherry blossom (1,335 mentions) being the most prominently perceived by the public. Additionally, 17 parks host flower-related cultural activities. 2) In terms of perception themes, the multiple functions of flower experiences (63.1%) are most prominent, followed by the aesthetic use of flower species (19%), the livelihood benefits of floral resources (14.4%), and the cultural value of floral attractions (3.5%). Public perception is more focused on functional aspects, such as photography and outings, while cultural values receive less attention. Among the three park types, the aesthetic use of flower species (22%) ranks second in historical parks, but is low in comprehensive and community parks, showing a negative correlation with livelihood benefits. In comprehensive parks, the multiple functions of flower experiences (68%) are more prominent, whereas in community parks, the livelihood benefits of floral resources (39%) are most noticeable. 3) In terms of the perception importance − satisfaction performance, advantageous flowers vary by park type, and the perception performance of various flower plant species is generally good. Lotus and chrysanthemums in historical parks are highly satisfactory, sunflowers in comprehensive parks are recognized, while chinese rose and peach blossom in community parks need to be upgraded. Besides, some flowers are highly rated while failing to attract enough attention, and limited resources need to be optimized. The perception theme performance of different types of parks varies greatly. The perception theme performance of historical famous parks is generally high. For comprehensive parks, the “multiple functions of flower experiences” stand out (68%), but the perception satisfaction is relatively low and in urgent need of enhancement. The overall satisfaction of each perception theme in community parks is relatively low.

Conclusion This research clarifies the primary species, thematic dimensions, and public performance evaluations related to floral landscapes across different types of urban parks in Beijing. It demonstrates that floral landscape design must be tailored to park functions, public usage patterns, and seasonal cycles to enhance user satisfaction and fulfill the goals of the Beijing Garden City initiative. The proposed perception evaluation model, not only aids local government and planners in refining floral landscape strategies but also provides a scalable methodological framework for evaluating similar projects in other cities pursuing green and livable urban transformations. Moreover, by highlighting gaps in satisfaction and underutilized floral resources, this research offers targeted recommendations for improving biodiversity, cultural engagement, and ecological value in urban landscape planning.