Objective  Urban public spaces are an essential carrier of everyday life, which can provide an interactive and integrative platform for urban social relationships at all levels. They embody the richness and diversity of urban life and are a crucial reflection of urban quality. It is necessary to clarify the formation mechanism of public space with high vitality. In recent years, a number of qualitative and quantitative studies have been conducted with a focus on the creation and influencing factors of public space vitality. However, with the reliance on urban infrastructure and information and communication technology, humanity has entered a new stage of information-driven society, leading to a new turn in spatial-temporal relationships. Although information flow is already an important intervention in everyday life in the mobile Internet era, it has received little attention. Therefore, this research explores the association between information flow and public space vitality using multi-source geographic big data.

Methods  Taking 125 urban parks in Shenzhen as examples, this research adopts multiple sets of urban big data to construct models to quantify the influence of information flow on the vitality of urban public spaces, and the data adopted include vector building and road data, remote sensing images, point of interest (POI) data, housing data, location-based service (LBS) data, and user-generated content (UGC) data. The UGC and LBS datasets are used to evaluate the information flows and vitality of urban public spaces, while the other datasets are used to calculate the covariates. Based on the collection of data and 316,929 UGCs with respect to the 125 sample parks in Shenzhen and the control of key covariates, this research quantifies the influence of information flow on public space vitality.

Results  Our research has four results: 1) using urban parks and UGC as an example, this research reveals the temporal and spatial characteristics of public space vitality and information flow; 2) the intervention of information flow has a significant positive effect on public space vitality; 3) compared to potential sentiments and exposure, the score of information flow has a stronger positive effect on public space vitality; 4) almost all geospatial information flows have a significant effect on public space vitality, while non-geospatial information flows do not have a significant effect. In detail, the research finds that the score, sentiment, and exposure of information flow in different public spaces are different. The score of 95% of sample public spaces falls within the range of 3.00~4.85, with a mean of 3.95, indicating an overall positive evaluation by users of public spaces. The distribution of sentiment values of information flow is similar to that of scores, with 95% of total sentiment values greater than 1 (i.e., positive sentiments), falling within the range of 1.16~1.91. The mean sentiment value is 1.70, indicating a generally high positive sentiment of information flow in public spaces. Compared to scores and sentiments, the exposure of information flow shows a significant difference, with 95% of the average exposure of each information falling within the range of 0~141.0, with a mean of 23.3. The time distribution of weekday and weekend vitality differs among sample public spaces. The peak hours of weekday public space vitality are from 09:00 to 13:00 and from 15:00 to 19:00. The peak hours of weekend public space vitality are similar to those on weekdays, but the fluctuation range of vitality is larger on weekends, and the peak value on weekends is significantly higher than that on weekdays. Moreover, the research reviews the significant association between information flow and public space vitality using regression models. The results show that the score of information flow has a positive and significant independent effect on public space vitality, and the average standardized effect of significant score variables in information flow is the highest among the three types of information flow (0.181). Similarly, the sentiment of information flow also has a significant effect on public space vitality, with an average standardized effect of 0.135 for significant sentiment variables in information flow. However, even though some exposure variables of information flow have a significant effect on public space vitality, the increase in adjusted R² is not significant (p > 0.050), implying that the promotion effect of exposure on public space vitality may be weaker than that of UGC scores and sentiment variables. Additionally, different types of information flow show significant differences: almost all variables of geolocation information flow have a positive and significant effect on public space vitality, while non-geolocation information flow has no variables with significant effects.

Conclusion  This research focuses on the information flow related to urban public spaces, and quantifies the influence of information flow on the creation of public space vitality. The research explores the significant effect of information flow and its heterogeneity, which will provide quantitative data support and strategies for the creation of public space vitality. In addition, the research further reflects on the relationship between human, technology, and the city to reconceptualize the contemporaneity of the creation of public space vitality.