Next Article in Journal
Multi-Objective Optimization Design of Geometric Parameters of Atrium in nZEB Based on Energy Consumption, Carbon Emission and Cost
Previous Article in Journal
Beyond the Smart or Resilient City: In Search of Sustainability in the Sojan Thirdspace
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Sustainability
Volume 15
Issue 1
10.3390/su15010146
sustainability-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Parks, Green Space, and Happiness: A Spatially Specific Sentiment Analysis Using Microblogs in Shanghai, China

by
Siqi Lai
* and
Brian Deal
Department of Landscape Architecture, University of Illinois at Urbana-Champaign, 611 Taft Drive, Champaign, IL 61820, USA
*
Author to whom correspondence should be addressed.
Sustainability 2023, 15(1), 146; https://doi.org/10.3390/su15010146
Submission received: 18 November 2022 / Revised: 18 December 2022 / Accepted: 20 December 2022 / Published: 22 December 2022
(This article belongs to the Section Health, Well-Being and Sustainability)
Browse Figures
Review Reports Versions Notes

Abstract

:
Green spaces, particularly urban parks, provide essential environmental, aesthetic, and recreational benefits to human health, well-being, and happiness. However, traditional forms of investigating people’s perceptions of urban parks, such as questionnaires and interviews, are often time- and resource-intensive and do not always yield results that are transferable across sites. In this study, spatially explicit geolocational information (Sina Weibo check-in data) was utilized to analyze expressions of happiness and well-being in urban parks in Shanghai, China. The results showed significant differences in reported happiness inside and outside urban parks in Shanghai over a 6-month period. Accessibility, naturalness factors, and the frequency of park visits were positively associated with happiness. There existed both commonalities and disparities in the results between residents and non-residents. These findings can provide decision makers and urban planners with a comprehensive and timely overview of urban park use so they can accurately identify park needs and improvements.

1. Introduction and Literature Review

There is considerable evidence in the literature that exposure to urban green spaces can be beneficial to a range of dimensions in human health [1,2,3]. Urban green spaces, particularly urban parks, have inherent environmental, aesthetic, and recreational attributes that contribute to human physical, mental, and social well-being [4,5], improve academic and job performance [6,7], and affect overall happiness [8]. Among these, however, the association between urban parks and happiness has received less critical attention.
A variety of terms are used in the cross-disciplinary literature on happiness, including happiness, subjective well-being, life satisfaction, experienced utility, and quality of life [9,10,11]. Various methods have also been employed to obtain and process well-being and happiness data. Self-reported measures (e.g., self-reported well-being) [12,13], questionnaires [14,15], and demographic surveys [16,17] are most commonly used to explore ‘happiness’. However, traditional methods of investigating people’s feelings are often time-consuming and resource-intensive, and they do not always produce results that can be transferred between sites. Smartphone apps and wearable sensors combined with GPS locational data have been used to obtain geospatially explicit data on feelings of well-being and happiness [18]. In most cases, however, this method is used in experimental settings. The limitation of this method lies in the fact that it is usually measured only once or twice on a limited number of participants.
In recent years, social media platforms have garnered interest in obtaining geospatially explicit and socially relevant data sets [19,20,21,22]. Platforms such as Twitter and Weibo have become ubiquitous and important for social networking and content sharing [23,24]. The widespread use of these platforms has enabled researchers to obtain large-scale and long-term databases on socially subjective information quickly and inexpensively [25]. Advanced artificial intelligence and deep learning techniques have helped magnify and improve on these advantages improved by advancing an ability to analyze social media text and conduct socially relevant sentiment analysis.
This paper uses the term, ‘sentiment analysis’ [26] to describe the process of data crawling and text analytics used to discern feelings and emotive content in a specific social media post. In the following, we describe a sentiment analysis of a large number of microblogs (Sina Weibo social media posts) with spatially explicit geolocational information to explore the effects of urban parks on expressions of well-being and happiness. This study uses the urban park system in Shanghai, China as the study site to address the following questions: (1) Do expressions of happiness and well-being made on social media posts increase in or around urban parks? (2) Does access to parks affect these expressions? (3) Does park ecosystem service availability (i.e., higher green coverage and/or the presence of water features) increase expressions of well-being and happiness? (4) What is the relationship between park visit frequency and expressions of happiness?
Our questions are addressed in the following format. The Introduction (Section 1) is followed by a review of the literature on supporting theories and studies connecting urban green space, emotive responses of well-being and happiness, and the use of social media data in sentiment analysis. Section 2 outlines our methods including study site, data analysis, variables, and measures. The results of our model analysis are described in Section 3, with a discussion and conclusions drawn in Section 4, including weaknesses of the study and the next steps. After the Section 4 is a description of the references cited. In general, we hope our analysis can provide planning guidance and design recommendations for urban parks.
Following is a review of the literature on theories and studies connecting urban green space and emotive responses to well-being and happiness.

1.1. Supporting Theories

The impacts of urban parks on happiness might largely be explained by some combination of five theoretical constructs: attention restoration [27], stress reduction [28], prospect-refuge [29], information processing [30], visual and acoustic stimuli [31].
(1)
Attention restoration theory posits that our top-down attention, which becomes depleted through mentally demanding tasks associated with everyday life, can be restored through exposure to natural settings [3,27,32]. Natural visual and acoustic elements, such as trees, water, tweeting birds, and breeze, are softly fascinating. These stimuli engage our bottom-up attention, thus allowing our top-down attention a chance to recover from the mental fatigue associated with modern life [27,33,34].
(2)
Stress reduction theory posits that exposure to nature promotes recovery from stress because positive mental responses to unthreatening natural settings and elements are deeply rooted in our genes through millions of years of evolution [28]. This psychological response is immediate, unconscious, and spontaneous, and it is accompanied by increased positive feelings [28,35,36].
(3)
The prospect-refuge theory [29] suggests that humans often experience unconscious positive mood responses to safe and fertile environments and experience negative mood responses to dangerous or barren settings. Prospect-refuge Theory has a great deal of overlap with the biophilia hypothesis, which posits humans possess an innate tendency to seek and prefer connections with nature [37,38].
(4)
The theory of information processing in landscape settings posits that making efficient and accurate comprehension of an environment is a foundational capacity that supported our ancestors’ survival and prosperity [30,39]. There is ample evidence that humans make rapid mental responses to landscapes and that these responses include changes in mood states [40]. Change in mood states was likely critical to facilitate our species’ ability to make immediate assessments of landscapes that might have posed resources or threats—or a combination of resources and threats [30,41,42].
(5)
Although a combination of all five senses is involved in human perception of any environment, it has been argued that the vast majority of information about landscapes is acquired by seeing and hearing [31]. This suggests that visual and acoustic stimuli are critical sources of information for connecting people to their surroundings [43]. In addition, these stimuli are linked, i.e., visual information can affect acoustic perception and acoustic information can affect visual perception [43,44,45]. The McGurk effect [46] for example, describes the incongruity experienced when well-known auditory and visual elements are mismatched. The effect is reported to be particularly salient when the quality of auditory information is poor [47]. Wang and Zhao (2019) [48] explore the effects of auditory-visual combinations on aesthetic preference, quantifying the advantage of soundscape and visual landscape congruence. From this perspective, a combination of visual and acoustic landscape variables might have varying effects on mood states [49].

1.2. Urban Parks and Happiness

The mechanisms linking happiness to urban parks have been the object of a wide range of scholarly research. For example, [50], and others [51,52,53] believe that urban parks can promote happiness by building adaptive capacity and restoring existing capacity. Ref. [52] summarize the mechanism of urban parks promoting happiness as harm reduction, resilience, and capacity building. Based on the existing research, the role of urban parks in promoting residents’ happiness can be divided into objective use behavior [54] and subjective perception behavior [55]. The former mainly encourages physical activity and ecosystem services, and the latter mainly relieves mental stress and promotes social interaction.
Urban parks are likely to provide a safe, accessible, and attractive setting in which to conduct physical activity [56,57]. There is even evidence (although inconsistent) that suggests that physical activity performed in urban parks produces greater psychological and physiological benefits than physical activity performed in other settings [58,59,60]. However, most studies to date, with only a few exceptions [58,61,62,63,64], have considered only the amount (duration, intensity) of physical activity conducted, and not whether the activity was performed in a greenspace or another setting. This limitation might at least partially explain the inconsistent evidence on the association between urban parks and “overall” physical activity levels [65,66].
Studies that have used mediation analysis to investigate whether physical activity lies on the causal pathway between urban parks and happiness have yielded mixed findings: some studies observed an indirect effect [64,67,68] while others did not [69,70,71,72]. What appears certain is that the sole presence of urban parks does not necessarily imply their use. In particular, not all urban parks are attractive for physical activity due to characteristics such as size and available facilities. For example, previous work has reported larger urban parks with well-maintained paths are likely to be more attractive to adults for physical activity than smaller “pocket parks”, which may be more attractive for more sedentary forms of recreation [73]. Differences in the urban park exposure indicators used [61] and types of urban parks included in the analysis could be further explanations for the existing inconsistent epidemiological findings.

1.3. Social Media Data and Sentiment Analysis

Data derived from social media platforms has been criticized in the literature for a lack of representativeness. Some argue that the demographic characteristics of social media users (age, gender, and ethnicity) are limited and may distort the resulting analysis [8]. Other studies argue, however, that generalizable results can be usefully extracted from geographically located social media data if the study is limited to metropolitan areas with high concentrations of smartphone users [74]. Unlike traditional data-gathering methods on greenspace perceptions (surveys, questionnaires, and interviews) that have a limited number of sampling units and a restricted spatio-temporal range [75], sentiment analysis derived from social media data can produce large amounts of information on social relationships and daily life in a cost-effective and timely way [76].
Sina Weibo is one such social media platform capable of producing large, easily retrievable, data sets on specific geographies. It is a self-described “microblogging” app and website based in China with more than 500 million users, considered the county’s most popular social media platform [77]. Big data sets from the platform can be delivered in an aggregated form rather than at an individual level, preserving personal privacy [78]. The platform has been found to be a useful source of data in large urban areas with many Weibo users [79].
Sentiment analysis is the process of understanding attitudes (positive, neutral, or negative) on a specific subject or topic expressed in social media posts [80]. It generally has two components: extracting emotional expressions (sentiment) and calculating their veracity [81]. Machine learning, lexicon-based, and hybrid approaches are the methodologies of choice for most current sentiment analyses [82,83,84]. Machine learning (ML) utilizes machine learning algorithms and linguistic features to conduct sentiment classification. Lexicon-based approaches rely on sentiment lexicons, which represent words and phrases commonly used to express positive and negative emotions [85]. A hybrid approach combines machine learning and lexicon-based approaches for a more finely tuned scaled analysis, although machine learning remains the most widely used approach [86].
The main advantage of the ML approach is that it can learn domain-specific patterns that improve classification results [87]. The downside is that it often requires large training datasets to achieve good results [88]. In this study, the volume of our datasets met the ML training requirement and was used in our analysis.
A subset of the ML approach is ‘deep learning’ (DL) or ‘deep structured learning’. DL uses artificial neural networks (ANNs) to create multiple layers in the network to progressively extract higher-level features from the data [89]. This approach allows for more complex models to be trained on a wider range of data sources [90]. DL includes many neural network models, including a recurrent neural network (RNN) [91]. No pre-defined features are needed in an RNN, as they can derive and remember long sequences of complex data interactions on their own [92]. A shortcoming of the RNN is that data gradients tend to get washed out over long time steps [93]. In response, [94] introduced a long-short term memory (LSTM) component of an RNN to capture and store information on data differences and gradients that are produced in intermittent time steps. Sentiment analysis using this type of architecture (DL RNN with LSTM) has become increasingly popular among researchers [95]. In this study, a ML approach using the DL RNN process with LSTM is used to analyze big data sets generated from Sina Weibo microblogs to measure sentiment among park users in Shanghai China.

2. Materials and Methods

Urban parks in Shanghai were selected for the study. This city was chosen because it is densely populated with a large volume of available data. The study area covers twelve districts in Shanghai, including Baoshan, Jiading, Jingan, Yangpu, Hongkou, Putuo, Changning, Huangpu, Xuhui, Songjiang, Minhang, and Pudong New districts. Parks with a low total number of microblog check-in records, a low total number of users, and poor data quality were excluded. Community parks and street parks were excluded, considering the area and service scope of the parks. In all, 39 parks were identified as sample parks for this study, including 22 comprehensive parks and 17 theme parks (Figure 1).

2.1. Subsection Data Sources and Processing

2.1.1. Sina Weibo Data

The social media site Sina Weibo, often called “Chinese Twitter”, was launched in August 2009 and has quickly become one of the most popular websites in China [77]. It offers a “check-in” feature where users share their microblogs and real-time location information. According to previous studies that used open crowdsourced data, check-in data derived from Sina Weibo is dependable and verifiable. It has also been used in previous studies on sentiment in relation to green spaces [77] As previously noted, this data makes it possible to measure sentiment without conducting expensive and time-consuming field investigations. We use Sina Weibo check-in data in this study to estimate the sentiment of Shanghai park visitors.
For 181 days between 1 January 2021 and 30 June 2021, Sina Weibo data were collected by web crawlers using the Sina Weibo application-programming interface (API). This study sets the characteristics of data capture to include: text, date, ID, name, and location. Three steps for the quantification of sentiment scores were established. (1) Data acquisition and cleaning, (2) text feature extraction, and (3) training sentiment models and classification. About 640,000 pieces of raw data were initially obtained. Of these, 570,000 were found useable after cleaning (duplicate data, ads, polls, image sharing, @someone, retweets, and other invalid texts were eliminated). Emojis were transformed into corresponding text using ‘emojiswitch’ library [96]. Word2Vec (a technique for natural language processing) was used to transform the text into numerical models so that a sentiment classification model could process the text data. Text sentiment analysis was then performed based on the DL RNN LSTM (Long short-term memory) model using Pytorch (an open-source machine learning framework) (see Figure 2).

2.1.2. Other Basic Data

This study collected data regarding travel time to Shanghai parks in three modes (walking, biking, and driving), the attributes of urban parks, and population. The path planning interface in Baidu Maps API (application programming interface) was used to find the shortest travel time from each residential area to the park. The basic attributes of parks (name, size, and type classifications) were collected from the Shanghai Star Park List (2020) published by Shanghai Landscaping & City Appearance Administrative Bureau. The population data were provided by the Institute of Geographical Sciences and Resources, Chinese Academy of Sciences and the time of the data is 2020. The spatial unit of population data is subdistrict.

2.2. Analytical Framework and Study Design

Structural equation modeling (SEM) is a method for developing, estimating, and testing causal models that contain both observable explicit variables and unobservable latent variables [97]. This study used maximum likelihood estimation to solve the structural equation model.
Taking Shanghai as a case study area, the structural equation model was elaborated and consists of a measurement model and a structural model (Figure 3). In which, the average sentiment score and frequency of park visiting were selected as endogenous variables, while the accessibility and natural degree of the parks were extracted according to the former studies as exogenous variables. In addition, the age of the study objects was used as a control variable. IBM SPSS Amos 28 Graphics was used to analyze the association between expressed happiness and various factors. ArcGIS Pro and Python (an interpreted, high-level programming language) were used for mapping the results. The following five hypotheses were proposed to verify the relationship between expressed happiness and urban park characteristics:
Hypothesis 1.
The greater the accessibility of the park, the happier people who have access to it.
Hypothesis 2.
The greater the natural degree of the park, the happier people who have access to it.
Hypothesis 3.
The more often people visit the park, the happier people who have access to it.
Hypothesis 4.
There are disparities in expressed happiness among different groups (residents and non-residents, female and male).

Variable Settings and Measures

The dependent variable for this study is “Expressed happiness”. The expressed happiness indicator is a measure of people’s sentiment level inside and outside of an urban park during a certain period and is expressed as the mean of the whole sentiment score. This value ranges from 0 to 1 and the larger the value of the indicator, the stronger the expressed happiness. In this study, the Weibo posts inside parks are about 20 K; the Weibo posts outside parks are about 620 K. We combined the Weibo posts inside and outside parks to represent the comprehensive individual sentiment level. Therefore, the large difference between the two data quantities may not affect the results on expressed well-being.
The mediating variable for this study is the “Frequency of park visits”. This variable has been widely applied in health studies due to its strong association with positive health outcomes [12,98,99]. For example, the research found that park use reduces stress levels [100]. As shown in Table 1, the distribution of frequency of park visit data was divided into three classes based on the natural breaks classification (Jenks) method [101]. In three classes, “0–1 times/month” (89.08%) and “2–4 times/month” (9.92%) had the largest number of people. This variable is a measure of people’s frequency of visiting the urban parks and is expressed as the ratio of the total times of visits to the amount of time (month). In this case, study, six months was used as the research period. In addition to this mediating variable, two socioeconomic factors (i.e., age and whether resident) were also included as control variables in the analysis.
The independent variables for this study are the characteristics of the urban park’s features. Based on the literature, an environmental factor index system that comprises accessibility and the natural degree of urban parks has been constructed (Table 2).
(1)
Accessibility
Accessibility refers to the convenience of reaching urban parks and is a common indicator used in landscape research. The commonly used measurement methods for accessibility include the proportional method, the coverage method, the closest distance method, the gravity model method, and the Gaussian two-step floating catchment area method. Among these methods, the Gaussian two-step floating catchment area method (G2SFCA) [102] considers both supply and demand factors and can comprehensively and easily calculate the accessibility of park green space. The travel cost in G2SFCA can be divided into distance cost and time cost. The traditional G2SFCA is based on distance cost. In recent research, time cost was used to improve the traditional one. More and more studies directly acquire time cost data for various travel modes through web mapping API (Application Program Interface) [103,104,105,106,107]. The advantages of using the time cost from the map API include it uses the latest road network and takes into account traffic congestion. In addition, it considers the different entrances at the origin and the destination (e.g., large-sized parks), which is a major factor affecting the time cost of the walking and biking mode [108]. Therefore, it can better reflect the actual connotation of accessibility [109].
This paper was carried out under the policy background of the “Shanghai 15 min walk life circle”, focusing on the accessibility of parks under walking conditions within 15 min. In this paper, the time cost was included in the accessibility calculation model:
A k T = ∑ k ∈ t k j ≤ t 0 G t k j , t 0 P j ∑ k ∈ t k j ≤ t 0 G t k j , t 0 × S k
G t k j , t 0 = e − 1 2 × t k j t 0 2 − e − 1 2 1 − e − 1 2     t k j ≤ t 0 0             t k j > t 0
where A k T is the accessibility of the park based on time cost, and the larger the calculated A k T , the better the accessibility of the park k ; t k j is the commuting time between origin j and park k , and the spatial unit of origin j is subdistrict; t 0 represents the threshold of travel time, which is set to 15 min; P j is the size of demand at origin j , measured in population; G t k j , t 0 is a Gaussian function; and S k represents the supply scale of park k within the search range, measured by its area.
(2)
Naturalness
A substantial body of research articulates the multiple contributions of green spaces, including fewer symptoms of depression [110] and lower levels of self-reported [111] and biologically measures stress [112]. Therefore, greenness coverage and water coverage were used to evaluate the naturalness of parks. The “Classification Wizard” (an image classifier based on deep learning) of ArcGIS Pro was used to classify satellite images into four classes: tree, lawn, water, and developed area. The ArcGIS Pro classification wizard classification employs the support vector machine (SVM) algorithm to analyze the remote sensing images. The precision of satellite image classification is over 96% [113]. Then, the area of greenness or water was used to calculate these variables.

2.3. Reliability and Validity Testing

To ensure the validity of model fit evaluation and hypothesis testing, it is necessary to check the reliability of variable measurements. Cronbach’s coefficient alpha (CCA) is used to test the reliability of the data. It is generally believed that data with a reliability coefficient above 0.6 is considered to have good reliability. The Cronbach reliability coefficients of the latent variables are all above 0.6 (Table 3), and the overall reliability of the data is also above 0.7, so we can consider the data to have good reliability.
Composite reliability, CR can be used as a reliability coefficient for a measurement instrument. A high-reliability coefficient indicates that the indicators are internally consistent. It is generally believed that data with a CR above 0.70 is considered to have good validity. As shown in Table 3, the use of observed variables to measure latent variables in this study was appropriate, and the measurement model had good internal consistency.
In addition, the discriminant validity of the measurement model was a test of the degree to which the latent variables are distinguished from each other, which could be determined by comparing the square root of the average variance extracted (AVE) of the latent variables with the size of the correlation coefficient between the latent variables. The AVE indicator should be at least 0.5. The model in this study had relatively good discriminant validity.

3. Results

Figure 4 showed significant differences in reported happiness inside and outside urban parks. Generally, the results showed that people who have access to urban park services express more feelings of happiness. The sentiment score inside parks starts to rise in February when the Spring festival starts. Based on text contents, the potential reason for the relatively low sentiment score in January may be the low temperature. Similarly, the fluctuation during May and June may also be caused by the high temperature.

Results of Structural Equation Model

The results (Figure 5) showed that both accessibility and naturalness factors were positively associated with the frequency of park visits and had an indirect effect on happiness (sentiment). Additionally, the coefficient value of accessibility was higher than the natural degree (greenness coverage and water coverage) (Table 4), which indicated that accessibility had a more positive effect on expressed happiness than naturalness. However, the direct effects of accessibility and naturalness on happiness were relatively weak because the values of the coefficient were relatively low. The frequency of park visits was positively associated with expressed happiness, and the direct effect value was 0.611. The frequency of park visits may mediate between urban parks and expressed happiness.
When comparing the results between residents and non-residents (Figure 6), there existed both commonalities and disparities. The residents and non-residents are distinguished by their location from their personal information. For both groups, the factors ‘age’ and ‘frequency of park visit’ had the highest values of coefficient. This result might indicate that people who are older and have a relatively high frequency of park visits tend to have a higher level of happiness. As for differences, ‘naturalness’ and ‘accessibility’ had a greater impact on the happiness of non-residents than residents. In addition, for residents, accessibility was more important than naturalness, while for non-residents, these two factors had the same relationship, the degree of difference between the two was smaller (resident: 0.097 − 0.065 = 0.032; non-resident: 0.129 − 0.103 = 0.026). The differences in results between genders were relatively small (Figure 7). For females, the greenness coverage was more important than for males while the water coverage was the opposite.

4. Discussion and Conclusions

The results showed significant differences in reported happiness inside and outside urban parks in Shanghai over a 6-month period. This study included the sentiment of people both inside and those with good accessibility to urban parks and their services. Previous studies on greenness and parks typically focused on age or ethnic groups and ignored the fact that not all people have good access to green services [2,114,115].
In exploring the factors associated with expressed happiness, there was a much higher coefficient of influence of park visits on happiness than other influencing factors (Table 5). This result had some discrepancies with previous studies. Several studies have sought to identify “features” and “characteristics” of urban parks that are linked with happiness, such as size, aesthetic appeal, and facilities [116,117]. However, this approach makes an assumption of a direct correlation between the features of an urban park and happiness. In reality, the relationship is more likely to be complex, multifactorial, and prone to considerable confounding [118]. Happiness is more likely to be directly linked to the frequency of park visits and activities that are undertaken in urban parks (social interaction, cultural activities, rest, and restitution). In other words, the result indicated that the mental health benefits accrued are a result of the use of urban parks and not just from the presence (or absence) of a park.
If the functionality of urban parks can be linked to people’s expressed happiness, different uses of urban parks should yield different kinds and degrees of mental health benefits. For example, social benefits are predicated on social contact happening in an urban park. That urban park may have to be situated in the right locality, carry some social meaning for the neighborhood, and its user groups permit or facilitate social interactions to occur. Consequently, modifications of the park in terms of the physical environment (e.g., accessibility and naturalness) alone may be insufficient to promote activities if there are significant psychosocial processes at play. These social ties, social networks, and social interactions may differ quite considerably between affluent and more deprived neighborhoods. Therefore, concomitant modifications of the social environment may be required.
In addition, based on the disparities between residents and non-residents, female and male, there is also a need to identify what mental health benefits are sought, what activities in urban parks contribute to these benefits, and in turn, identify what features of an urban park would encourage such activities.

4.1. Conclusions

This study demonstrates a method for analyzing the spatiotemporal distribution of expressed happiness involving social media data. The test case was urban parks in Shanghai, China. This model can also be used to evaluate happiness in other cities since social media data is relatively accessible, although care should be taken to choose data sets large enough to overcome some of the limitations of using social media-derived data sets. In addition, during the COVID-19 pandemic, these findings can assist in the planning and design of urban parks that are health-oriented.
In the structural equation model, all hypotheses were supported. This finding is meaningful because it can provide decision-makers and urban planners a comprehensive and timely overview of urban park use so they can accurately identify park needs and improvements. These findings can also provide urban planners a reference for enhancing the overall urban park neighborhood environment.

4.2. Limitations and Future Steps

There are three main limitations in this study. First, this study was conducted in Shanghai, a city with a high level of social and economic development, so the applicability of the findings of this study to other cities has yet to be verified. Second, Sina Weibo data has limitations: (1) the generation of positioning data depends on active operations by the user, so there are cases where data sampling is lost. (2) The social media information may not be accurate. Future studies should combine the traditional data (questionnaire and interview) with social media data so that we can gain a more accurate and comprehensive understanding of residents’ sentiments. Third, other influence paths (reducing harm and building social connections) were ignored, and they should be considered in future work.

Author Contributions

Conceptualization, S.L. and B.D.; methodology, S.L. and B.D.; software, S.L.; validation, S.L. and B.D.; formal analysis, S.L.; investigation, S.L.; resources, S.L. and B.D.; data curation, S.L.; writing—original draft preparation, S.L.; writing—review and editing, B.D.; visualization, S.L.; supervision, B.D.; project administration, B.D.; funding acquisition, B.D. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Not applicable.

Acknowledgments

We thank Candidate Yifan Zhu from the Computer Science department at UIUC for providing advice on data processing. We appreciate the constructive suggestions from the anonymous reviewers.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Bratman, G.N.; Anderson, C.B.; Berman, M.G.; Cochran, B.; de Vries, S.; Flanders, J.; Folke, C.; Frumkin, H.; Gross, J.J.; Hartig, T.; et al. Nature and Mental Health: An Ecosystem Service Perspective. Sci. Adv. 2019, 5, eaax0903. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  2. Enssle, F.; Kabisch, N. Urban Green Spaces for the Social Interaction, Health and Well-Being of Older People— An Integrated View of Urban Ecosystem Services and Socio-Environmental Justice. Environ. Sci. Policy 2020, 109, 36–44. [Google Scholar] [CrossRef]
  3. Sullivan, W.C.; Chang, C.Y. Landscapes and Human Health. Int. J. Environ. Res. Public Health 2017, 14, 1212. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  4. Frumkin, H.; Bratman, G.N.; Breslow, S.J.; Cochran, B.; Kahn, P.H.; Lawler, J.J.; Levin, P.S.; Tandon, P.S.; Varanasi, U.; Wolf, K.L.; et al. Nature Contact and Human Health: A Research Agenda. Environ. Health Perspect. 2017, 125, 075001. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  5. Villeneuve, P.J.; Jerrett, M.; Su, J.G.; Burnett, R.T.; Chen, H.; Wheeler, A.J.; Goldberg, M.S. A Cohort Study Relating Urban Green Space with Mortality in Ontario, Canada. Environ. Res. 2012, 115, 51–58. [Google Scholar] [CrossRef]
  6. Li, D.; Chiang, Y.C.; Sang, H.; Sullivan, W.C. Beyond the School Grounds: Links between Density of Tree Cover in School Surroundings and High School Academic Performance. Urban For. Urban Green. 2019, 38, 42–53. [Google Scholar] [CrossRef]
  7. Gascon, M.; Triguero-Mas, M.; Martínez, D.; Dadvand, P.; Rojas-Rueda, D.; Plasència, A.; Nieuwenhuijsen, M.J. Residential Green Spaces and Mortality: A Systematic Review. Environ. Int. 2016, 86, 60–67. [Google Scholar] [CrossRef] [Green Version]
  8. Cheng, Y.; Zhang, J.; Wei, W.; Zhao, B. Effects of Urban Parks on Residents’ Expressed Happiness before and during the COVID-19 Pandemic. Landsc. Urban Plan. 2021, 212, 104118. [Google Scholar] [CrossRef]
  9. De Vries, S.; Nieuwenhuizen, W.; Farjon, H.; van Hinsberg, A.; Dirkx, J. In Which Natural Environments Are People Happiest? Large-Scale Experience Sampling in the Netherlands. Landsc. Urban Plan. 2021, 205, 103972. [Google Scholar] [CrossRef]
  10. MacKerron, G.; Mourato, S. Happiness Is Greater in Natural Environments. Glob. Environ. Chang. 2013, 23, 992–1000. [Google Scholar] [CrossRef]
  11. Mackerron, G. Happiness and Environmental Quality. Ph.D. Thesis, London School of Economics and Political Science, London, UK, 2011. [Google Scholar]
  12. Van Herzele, A.; de Vries, S. Linking Green Space to Health: A Comparative Study of Two Urban Neighbourhoods in Ghent, Belgium. Popul. Environ. 2012, 34, 171–193. [Google Scholar] [CrossRef]
  13. Aminzadeh, K.; Denny, S.; Utter, J.; Milfont, T.L.; Ameratunga, S.; Teevale, T.; Clark, T. Neighbourhood Social Capital and Adolescent Self-Reported Wellbeing in New Zealand: A Multilevel Analysis. Soc. Sci. Med. 2013, 84, 13–21. [Google Scholar] [CrossRef]
  14. White, M.P.; Alcock, I.; Wheeler, B.W.; Depledge, M.H. Would You Be Happier Living in a Greener Urban Area? A Fixed-Effects Analysis of Panel Data. Psychol. Sci. 2013, 24, 920–928. [Google Scholar] [CrossRef]
  15. Nielsen, K.; Abildgaard, J.S.; Daniels, K. Putting Context into Organizational Intervention Design: Using Tailored Questionnaires to Measure Initiatives for Worker Well-Being. Hum. Relat. 2014, 67, 1537–1560. [Google Scholar] [CrossRef] [Green Version]
  16. Bertram, C.; Rehdanz, K. The Role of Urban Green Space for Human Well-Being. Ecol. Econ. 2015, 120, 139–152. [Google Scholar] [CrossRef] [Green Version]
  17. Blasi, E.; Nucera, M.; Cicatiello, C.; Franco, S. Socio-Demographic Components of Eudaimonic Well-Being: A Survey in an Italian Province. Soc. Indic. Res. 2013, 113, 451–470. [Google Scholar] [CrossRef]
  18. Benita, F.; Bansal, G.; Tunçer, B. Public Spaces and Happiness: Evidence from a Large-Scale Field Experiment. Health Place 2019, 56, 9–18. [Google Scholar] [CrossRef]
  19. Shan, S.; Peng, J.; Wei, Y. Environmental Sustainability Assessment 2.0: The Value of Social Media Data for Determining the Emotional Responses of People to River Pollution—A Case Study of Weibo (Chinese Twitter). Socioecon. Plann. Sci. 2021, 75, 100868. [Google Scholar] [CrossRef]
  20. Ghahramani, M.; Galle, N.J.; Ratti, C.; Pilla, F. Tales of a City: Sentiment Analysis of Urban Green Space in Dublin. Cities 2021, 119, 103395. [Google Scholar] [CrossRef]
  21. Seresinhe, C.I.; Preis, T.; MacKerron, G.; Moat, H.S. Happiness Is Greater in More Scenic Locations. Sci. Rep. 2019, 9, 4498. [Google Scholar] [CrossRef]
  22. Wang, Z.; Zhu, Z.; Xu, M.; Qureshi, S. Fine-Grained Assessment of Greenspace Satisfaction at Regional Scale Using Content Analysis of Social Media and Machine Learning. Sci. Total Environ. 2021, 776, 145908. [Google Scholar] [CrossRef] [PubMed]
  23. Hughes, D.J.; Rowe, M.; Batey, M.; Lee, A. A Tale of Two Sites: Twitter vs. Facebook and the Personality Predictors of Social Media Usage. Comput. Human Behav. 2012, 28, 561–569. [Google Scholar] [CrossRef] [Green Version]
  24. Gao, Y.; Chen, Y.; Mu, L.; Gong, S.; Zhang, P.; Liu, Y. Measuring Urban Sentiments from Social Media Data: A Dual-Polarity Metric Approach. J. Geogr. Syst. 2022, 24, 199–221. [Google Scholar] [CrossRef]
  25. Roberts, H.; Sadler, J.; Chapman, L. The Value of Twitter Data for Determining the Emotional Responses of People to Urban Green Spaces: A Case Study and Critical Evaluation. Urban Stud. 2019, 56, 818–835. [Google Scholar] [CrossRef] [Green Version]
  26. Plunz, R.A.; Zhou, Y.; Carrasco Vintimilla, M.I.; Mckeown, K.; Yu, T.; Uguccioni, L.; Sutto, M.P. Twitter Sentiment in New York City Parks as Measure of Well-Being. Landsc. Urban Plan. 2019, 189, 235–246. [Google Scholar] [CrossRef]
  27. Kaplan, S. The Restorative Benefits of Nature: Toward an Integrative Framework. J. Environ. Psychol. 1995, 15, 169–182. [Google Scholar] [CrossRef]
  28. Ulrich, R.S.; Simons, R.F.; Losito, B.D.; Fiorito, E.; Miles, M.A.; Zelson, M. Stress Recovery during Exposure to Natural and Urban Environments. J. Environ. Psychol. 1991, 11, 201–230. [Google Scholar] [CrossRef]
  29. Appleton, J. The Experience of Landscape; Wiley Chichester: Chichester, UK, 1975. [Google Scholar]
  30. Kaplan, S. Aesthetics, Affect, and Cognition: Environmental Preference from an Evolutionary Perspective. Environ. Behav. 1987, 19, 3–32. [Google Scholar] [CrossRef] [Green Version]
  31. Treichler, D.G. Are You Missing the Boat in Training Aids. Film. AV Commun. 1967, 1, 14–16. [Google Scholar]
  32. Sullivan, W.C.; Li, D. Nature and Attention. In Nature and Psychology; Springer: Berlin/Heidelberg, Germany, 2021; pp. 7–30. [Google Scholar]
  33. Ohly, H.; White, M.P.; Wheeler, B.W.; Bethel, A.; Ukoumunne, O.C.; Nikolaou, V.; Garside, R. Attention Restoration Theory: A Systematic Review of the Attention Restoration Potential of Exposure to Natural Environments. J. Toxicol. Environ. Health Part B 2016, 19, 305–343. [Google Scholar] [CrossRef] [Green Version]
  34. Stevenson, M.P.; Schilhab, T.; Bentsen, P. Attention Restoration Theory II: A Systematic Review to Clarify Attention Processes Affected by Exposure to Natural Environments. J. Toxicol. Environ. Health Part B 2018, 21, 227–268. [Google Scholar] [CrossRef]
  35. Andrade, C.C.; Devlin, A.S. Stress Reduction in the Hospital Room: Applying Ulrich’s Theory of Supportive Design. J. Environ. Psychol. 2015, 41, 125–134. [Google Scholar] [CrossRef] [Green Version]
  36. Jiang, B.; He, J.; Chen, J.; Larsen, L.; Wang, H. Perceived Green at Speed: A Simulated Driving Experiment Raises New Questions for Attention Restoration Theory and Stress Reduction Theory. Environ. Behav. 2021, 53, 296–335. [Google Scholar] [CrossRef]
  37. Wilson, E.O. Biophilia; Harvard University Press: Cambridge, MA, USA, 2021. [Google Scholar]
  38. Dosen, A.S.; Ostwald, M.J. Evidence for Prospect-Refuge Theory: A Meta-Analysis of the Findings of Environmental Preference Research. City Territ. Archit. 2016, 3, 4. [Google Scholar] [CrossRef] [Green Version]
  39. Kaplan, R.; Kaplan, S.; Ryan, R. With People in Mind: Design and Management of Everyday Nature; Island Press: Washinton, DC, USA, 1998. [Google Scholar]
  40. Kaplan, R.; Kaplan, S. The Experience of Nature: A Psychological Perspective; Cambridge University Press: Cambridge, UK, 1989. [Google Scholar]
  41. Opdam, P. Navigating the Space between Landscape Science and Collective Action for Sustainability: Identifying Key Factors in Information Processing. Landsc. Ecol. 2020, 35, 2629–2639. [Google Scholar] [CrossRef]
  42. Gimblett, H.R.; Itami, R.M.; Fitzgibbon, J.E. Mystery in an Information Processing Model of Landscape Preference. Landsc. J. 1985, 4, 87–95. [Google Scholar] [CrossRef]
  43. Preis, A.; Kociński, J.; Hafke-Dys, H.; Wrzosek, M. Audio-Visual Interactions in Environment Assessment. Sci. Total Environ. 2015, 523, 191–200. [Google Scholar] [CrossRef]
  44. Xu, X.; Wu, H. Audio-Visual Interactions Enhance Soundscape Perception in China’s Protected Areas. Urban For. Urban Green. 2021, 61, 127090. [Google Scholar] [CrossRef]
  45. Jeon, J.Y.; Jo, H.I. Effects of Audio-Visual Interactions on Soundscape and Landscape Perception and Their Influence on Satisfaction with the Urban Environment. Build. Environ. 2020, 169, 106544. [Google Scholar] [CrossRef]
  46. McGurk, H.; MacDonald, J. Hearing Lips and Seeing Voices. Nature 1976, 264, 746–748. [Google Scholar] [CrossRef]
  47. Massaro, D.W.; Cohen, M.M. Tests of Auditory–Visual Integration Efficiency within the Framework of the Fuzzy Logical Model of Perception. J. Acoust. Soc. Am. 2000, 108, 784–789. [Google Scholar] [CrossRef] [PubMed]
  48. Wang, R.; Zhao, J. A Good Sound in the Right Place: Exploring the Effects of Auditory-Visual Combinations on Aesthetic Preference. Urban For. Urban Green. 2019, 43, 126356. [Google Scholar] [CrossRef]
  49. Jiang, B.; Xu, W.; Ji, W.; Kim, G.; Pryor, M.; Sullivan, W.C. Impacts of Nature and Built Acoustic-Visual Environments on Human’s Multidimensional Mood States: A Cross-Continent Experiment. J. Environ. Psychol. 2021, 77, 101659. [Google Scholar] [CrossRef]
  50. Dzhambov, A.M.; Markevych, I.; Hartig, T.; Tilov, B.; Arabadzhiev, Z.; Stoyanov, D.; Gatseva, P.; Dimitrova, D.D. Multiple Pathways Link Urban Green- and Bluespace to Mental Health in Young Adults. Environ. Res. 2018, 166, 223–233. [Google Scholar] [CrossRef] [PubMed]
  51. Hartig, T.; Mitchell, R.; de Vries, S.; Frumkin, H. Nature and Health. Annu. Rev. Public Health 2014, 35, 207–228. [Google Scholar] [CrossRef] [Green Version]
  52. Markevych, I.; Schoierer, J.; Hartig, T.; Chudnovsky, A.; Hystad, P.; Dzhambov, A.M.; de Vries, S.; Triguero-Mas, M.; Brauer, M.; Nieuwenhuijsen, M.J.; et al. Exploring Pathways Linking Greenspace to Health: Theoretical and Methodological Guidance. Environ. Res. 2017, 158, 301–317. [Google Scholar] [CrossRef]
  53. Ikiugu, M.N.; Ciaravino, E.A. Psychosocial Conceptual Practice Models in Occupational Therapy: Building Adaptive Capability; Elsevier Health Sciences: Amsterdam, The Netherlands, 2007. [Google Scholar]
  54. Liu, H.; Li, F.; Li, J.; Zhang, Y. The Relationships between Urban Parks, Residents’ Physical Activity, and Mental Health Benefits: A Case Study from Beijing, China. J. Environ. Manag. 2017, 190, 223–230. [Google Scholar] [CrossRef]
  55. Kim, D.; Jin, J. Does Happiness Data Say Urban Parks Are Worth It? Landsc. Urban Plan. 2018, 178, 1–11. [Google Scholar] [CrossRef]
  56. Almanza, E.; Jerrett, M.; Dunton, G.; Seto, E.; Pentz, M.A. A Study of Community Design, Greenness, and Physical Activity in Children Using Satellite, GPS and Accelerometer Data. Health Place 2012, 18, 46–54. [Google Scholar] [CrossRef] [Green Version]
  57. Astell-Burt, T.; Mitchell, R.; Hartig, T. The Association between Green Space and Mental Health Varies across the Lifecourse. A Longitudinal Study. J. Epidemiol. Community Health 2014, 68, 578–583. [Google Scholar] [CrossRef]
  58. Mitchell, R. Is Physical Activity in Natural Environments Better for Mental Health than Physical Activity in Other Environments? Soc. Sci. Med. 2013, 91, 130–134. [Google Scholar] [CrossRef]
  59. Pretty, J.; Peacock, J.; Hine, R.; Sellens, M.; South, N.; Griffin, M. Green Exercise in the UK Countryside: Effects on Health and Psychological Well-Being, and Implications for Policy and Planning. J. Environ. Plan. Manag. 2007, 50, 211–231. [Google Scholar] [CrossRef]
  60. Thompson Coon, J.; Boddy, K.; Stein, K.; Whear, R.; Barton, J.; Depledge, M.H. Does Participating in Physical Activity in Outdoor Natural Environments Have a Greater Effect on Physical and Mental Wellbeing than Physical Activity Indoors? A Systematic Review. Environ. Sci. Technol. 2011, 45, 1761–1772. [Google Scholar] [CrossRef]
  61. Kaczynski, A.T.; Potwarka, L.R.; Smale, B.J.A.; Havitz, M.E. Association of Parkland Proximity with Neighborhood and Park-Based Physical Activity: Variations by Gender and Age. Leis. Sci. 2009, 31, 174–191. [Google Scholar] [CrossRef]
  62. Ord, K.; Mitchell, R.; Pearce, J. Is Level of Neighbourhood Green Space Associated with Physical Activity in Green Space? Int. J. Behav. Nutr. Phys. Act. 2013, 10, 1–8. [Google Scholar] [CrossRef] [Green Version]
  63. Schipperijn, J.; Bentsen, P.; Troelsen, J.; Toftager, M.; Stigsdotter, U.K. Associations between Physical Activity and Characteristics of Urban Green Space. Urban For. Urban Green. 2013, 12, 109–116. [Google Scholar] [CrossRef]
  64. De Vries, S.; van Dillen, S.M.; Groenewegen, P.P.; Spreeuwenberg, P. Streetscape Greenery and Health: Stress, Social Cohesion and Physical Activity as Mediators. Soc. Sci. Med. 2013, 94, 26–33. [Google Scholar] [CrossRef] [Green Version]
  65. Bancroft, C.; Joshi, S.; Rundle, A.; Hutson, M.; Chong, C.; Weiss, C.C.; Genkinger, J.; Neckerman, K.; Lovasi, G. Association of Proximity and Density of Parks and Objectively Measured Physical Activity in the United States: A Systematic Review. Soc. Sci. Med. 2015, 138, 22–30. [Google Scholar] [CrossRef]
  66. Lachowycz, K.; Jones, A.P. Greenspace and Obesity: A Systematic Review of the Evidence. Obes. Rev. 2011, 12, e183–e189. [Google Scholar] [CrossRef]
  67. McEachan, R.R.C.; Prady, S.L.; Smith, G.; Fairley, L.; Cabieses, B.; Gidlow, C.; Wright, J.; Dadvand, P.; van Gent, D.; Nieuwenhuijsen, M.J. The Association between Green Space and Depressive Symptoms in Pregnant Women: Moderating Roles of Socioeconomic Status and Physical Activity. J. Epidemiol. Community Health 2016, 70, 253–259. [Google Scholar] [CrossRef] [Green Version]
  68. Richardson, E.A.; Pearce, J.; Mitchell, R.; Kingham, S. Role of Physical Activity in the Relationship between Urban Green Space and Health. Public Health 2013, 127, 318–324. [Google Scholar] [CrossRef] [PubMed]
  69. Dadvand, P.; Bartoll, X.; Basagaña, X.; Dalmau-Bueno, A.; Martinez, D.; Ambros, A.; Cirach, M.; Triguero-Mas, M.; Gascon, M.; Borrell, C.; et al. Green Spaces and General Health: Roles of Mental Health Status, Social Support, and Physical Activity. Environ. Int. 2016, 91, 161–167. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  70. Astell-Burt, T.; Feng, X.; Kolt, G.S. Mental Health Benefits of Neighbourhood Green Space Are Stronger among Physically Active Adults in Middle-to-Older Age: Evidence from 260,061 Australians. Prev. Med. 2013, 57, 601–606. [Google Scholar] [CrossRef] [PubMed]
  71. Maas, J.; Verheij, R.A.; Spreeuwenberg, P.; Groenewegen, P.P. Physical Activity as a Possible Mechanism behind the Relationship between Green Space and Health: A Multilevel Analysis. BMC Public Health 2008, 8, 206. [Google Scholar] [CrossRef] [Green Version]
  72. Cohen, D.A.; Han, B.; Marsh, T.; Sturm, R. Quantifying the Contribution of Public Parks to Physical Activity and Health: Introducing SOPARC; RAND Corporation: Santa Monica, CA, USA, 2014. [Google Scholar]
  73. Giles-Corti, B.; Timperio, A.; Bull, F.; Pikora, T. Understanding Physical Activity Environmental Correlates: Increased Specificity for Ecological Models. Exerc. Sport Sci. Rev. 2005, 33, 175–181. [Google Scholar] [CrossRef]
  74. Liu, X.; Hu, W. Attention and Sentiment of Chinese Public toward Green Buildings Based on Sina Weibo. Sustain. Cities Soc. 2019, 44, 550–558. [Google Scholar] [CrossRef]
  75. Balram, S.; Dragićević, S. Attitudes toward Urban Green Spaces: Integrating Questionnaire Survey and Collaborative GIS Techniques to Improve Attitude Measurements. Landsc. Urban Plan. 2005, 71, 147–162. [Google Scholar] [CrossRef]
  76. Li, S.; Wang, Y.; Xue, J.; Zhao, N.; Zhu, T. The Impact of Covid-19 Epidemic Declaration on Psychological Consequences: A Study on Active Weibo Users. Int. J. Environ. Res. Public Health 2020, 17, 2032. [Google Scholar] [CrossRef] [Green Version]
  77. Zhu, J.; Xu, C. Sina Microblog Sentiment in Beijing City Parks as Measure of Demand for Urban Green Space during the COVID-19. Urban For. Urban Green. 2021, 58, 126913. [Google Scholar] [CrossRef]
  78. Han, X.; Wang, J.; Zhang, M.; Wang, X. Using Social Media to Mine and Analyze Public Opinion Related to COVID-19 in China. Int. J. Environ. Res. Public Health 2020, 17, 2788. [Google Scholar] [CrossRef] [Green Version]
  79. Fang, J.; Hu, J.; Shi, X.; Zhao, L. Assessing Disaster Impacts and Response Using Social Media Data in China: A Case Study of 2016 Wuhan Rainstorm. Int. J. Disaster Risk Reduct. 2019, 34, 275–282. [Google Scholar] [CrossRef]
  80. Nasukawa, T.; Yi, J. Sentiment Analysis: Capturing Favorability Using Natural Language Processing. In Proceedings of the 2nd International Conference on Knowledge Capture, Sanibel Island, FL, USA, 23–25 October 2003; pp. 70–77. [Google Scholar]
  81. Yi, J.; Nasukawa, T.; Bunescu, R.; Niblack, W. Sentiment Analyzer: Extracting Sentiments about a given Topic Using Natural Language Processing Techniques. In Proceedings of the Third IEEE international Conference on Data Mining, Melbourne, FL, USA, 22 November 2003; pp. 427–434. [Google Scholar]
  82. Thakkar, H.; Patel, D. Approaches for Sentiment Analysis on Twitter: A State-of-Art Study. arXiv 2015. [Google Scholar] [CrossRef]
  83. Sankar, H.; Subramaniyaswamy, V. Investigating Sentiment Analysis Using Machine Learning Approach. In Proceedings of the 2017 International Conference on Intelligent Sustainable Systems (ICISS), Palladam, India, 7–8 December 2017; pp. 87–92. [Google Scholar]
  84. Ligthart, A.; Catal, C.; Tekinerdogan, B. Systematic Reviews in Sentiment Analysis: A Tertiary Study. Artif. Intell. Rev. 2021, 54, 4997–5053. [Google Scholar] [CrossRef]
  85. Jurek, A.; Mulvenna, M.D.; Bi, Y. Improved Lexicon-Based Sentiment Analysis for Social Media Analytics. Secur. Inform. 2015, 4, 9. [Google Scholar] [CrossRef] [Green Version]
  86. Birjali, M.; Kasri, M.; Beni-Hssane, A. A Comprehensive Survey on Sentiment Analysis: Approaches, Challenges and Trends. Knowl. Based Syst. 2021, 226, 107134. [Google Scholar] [CrossRef]
  87. Agarwal, B.; Mittal, N. Prominent Feature Extraction for Sentiment Analysis; Springer: Berlin/Heidelberg, Germany, 2016. [Google Scholar]
  88. Yoo, G.; Nam, J. A Hybrid Approach to Sentiment Analysis Enhanced by Sentiment Lexicons and Polarity Shifting Devices. In Proceedings of the 13th Workshop on Asian Language Resources, Miyazaki, Japan, 7 May 2018; pp. 21–28. [Google Scholar]
  89. Vateekul, P.; Koomsubha, T. A Study of Sentiment Analysis Using Deep Learning Techniques on Thai Twitter Data. In Proceedings of the 2016 13th International Joint Conference on Computer Science and Software Engineering (JCSSE), Khon Kaen, Thailand, 13–15 July 2016; pp. 1–6. [Google Scholar]
  90. Zhang, L.; Wang, S.; Liu, B. Deep Learning for Sentiment Analysis: A Survey. Wiley Interdiscip. Rev. Data Min. Knowl. Discov. 2018, 8, e1253. [Google Scholar] [CrossRef] [Green Version]
  91. Li, W.; Qi, F.; Tang, M.; Yu, Z. Bidirectional LSTM with Self-Attention Mechanism and Multi-Channel Features for Sentiment Classification. Neurocomputing 2020, 387, 63–77. [Google Scholar] [CrossRef]
  92. Araque, O.; Corcuera-Platas, I.; Sánchez-Rada, J.F.; Iglesias, C.A. Enhancing Deep Learning Sentiment Analysis with Ensemble Techniques in Social Applications. Expert Syst. Appl. 2017, 77, 236–246. [Google Scholar] [CrossRef]
  93. Sharfuddin, A.A.; Tihami, M.N.; Islam, M.S. A Deep Recurrent Neural Network with Bilstm Model for Sentiment Classification. In Proceedings of the 2018 International Conference on Bangla Speech and Language Processing (ICBSLP), Sylhet, Bangladesh, 21–22 September 2018; pp. 1–4. [Google Scholar]
  94. Hochreiter, S.; Schmidhuber, J. Long Short-Term Memory. Neural Comput. 1997, 9, 1735–1780. [Google Scholar] [CrossRef]
  95. Yu, Y.; Si, X.; Hu, C.; Zhang, J. A Review of Recurrent Neural Networks: LSTM Cells and Network Architectures. Neural Comput. 2019, 31, 1235–1270. [Google Scholar] [CrossRef]
  96. Qin, J.; Huang, J. Use Pre-Trained Models and Multi-Classifier Voting Methods to Identify the Ironic Authors on Twitter. In Proceedings of the 13th Conference and Labs of the Evaluation Forum (CLEF), Bologna, Italy, 5–8 September 2022; pp. 1613–1673. [Google Scholar]
  97. Bollen, K.A. A New Incremental Fit Index for General Structural Equation Models. Sociol. Methods Res. 1989, 17, 303–316. [Google Scholar] [CrossRef]
  98. Nutsford, D.; Pearson, A.L.; Kingham, S. An Ecological Study Investigating the Association between Access to Urban Green Space and Mental Health. Public Health 2013, 127, 1005–1011. [Google Scholar] [CrossRef] [PubMed]
  99. Wood, L.; Hooper, P.; Foster, S.; Bull, F. Public Green Spaces and Positive Mental Health—Investigating the Relationship between Access, Quantity and Types of Parks and Mental Wellbeing. Health Place 2017, 48, 63–71. [Google Scholar] [CrossRef] [PubMed]
  100. Orsega-Smith, E.; Mowen, A.J.; Payne, L.L.; Godbey, G. The Interaction of Stress and Park Use on Psycho-Physiological Health in Older Adults. J. Leis. Res. 2004, 36, 232–256. [Google Scholar] [CrossRef]
  101. Chen, J.; Yang, S.; Li, H.; Zhang, B.; Lv, J. Research on Geographical Environment Unit Division Based on the Method of Natural Breaks (Jenks). Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci. 2013, 40, 47–50. [Google Scholar] [CrossRef] [Green Version]
  102. Hu, S.; Song, W.; Li, C.; Lu, J. A Multi-Mode Gaussian-Based Two-Step Floating Catchment Area Method for Measuring Accessibility of Urban Parks. Cities 2020, 105, 102815. [Google Scholar] [CrossRef]
  103. Zheng, Z.; Xia, H.; Ambinakudige, S.; Qin, Y.; Li, Y.; Xie, Z.; Zhang, L.; Gu, H. Spatial Accessibility to Hospitals Based on Web Mapping API: An Empirical Study in Kaifeng, China. Sustainability 2019, 11, 1160. [Google Scholar] [CrossRef] [Green Version]
  104. Xia, N.; Cheng, L.; Chen, S.; Wei, X.Y.; Zong, W.W.; Li, M.C. Accessibility Based on Gravity-Radiation Model and Google Maps API: A Case Study in Australia. J. Transp. Geogr. 2018, 72, 178–190. [Google Scholar] [CrossRef]
  105. Tao, Z.; Yao, Z.; Kong, H.; Duan, F.; Li, G. Spatial Accessibility to Healthcare Services in Shenzhen, China: Improving the Multi-Modal Two-Step Floating Catchment Area Method by Estimating Travel Time via Online Map APIs. BMC Health Serv. Res. 2018, 18, 345. [Google Scholar] [CrossRef] [Green Version]
  106. García-Albertos, P.; Picornell, M.; Salas-Olmedo, M.H.; Gutiérrez, J. Exploring the Potential of Mobile Phone Records and Online Route Planners for Dynamic Accessibility Analysis. Transp. Res. Part. A Policy Pract. 2019, 125, 294–307. [Google Scholar] [CrossRef]
  107. Tao, Z.; Cheng, Y. Modelling the Spatial Accessibility of the Elderly to Healthcare Services in Beijing, China. Environ. Plan. B Urban Anal. City Sci. 2019, 46, 1132–1147. [Google Scholar] [CrossRef]
  108. Qin, J.; Liu, Y.; Yi, D.; Sun, S.; Zhang, J. Spatial Accessibility Analysis of Parks with Multiple Entrances Based on Real-Time Travel: The Case Study in Beijing. Sustainability 2020, 12, 7618. [Google Scholar] [CrossRef]
  109. Gong, S.; Gao, Y.; Zhang, F.; Mu, L.; Kang, C.; Liu, Y. Evaluating Healthcare Resource Inequality in Beijing, China Based on an Improved Spatial Accessibility Measurement. Trans. GIS 2021, 25, 1504–1521. [Google Scholar] [CrossRef]
  110. Miles, R.; Coutts, C.; Mohamadi, A. Neighborhood Urban Form, Social Environment, and Depression. J. Urban Health 2012, 89, 1–18. [Google Scholar] [CrossRef] [Green Version]
  111. Den Berg, A.E.; Maas, J.; Verheij, R.A.; Groenewegen, P.P. Green Space as a Buffer between Stressful Life Events and Health. Soc. Sci. Med. 2010, 70, 1203–1210. [Google Scholar] [CrossRef] [Green Version]
  112. Thompson, C.W.; Roe, J.; Aspinall, P.; Mitchell, R.; Clow, A.; Miller, D. More Green Space Is Linked to Less Stress in Deprived Communities: Evidence from Salivary Cortisol Patterns. Landsc. Urban Plan. 2012, 105, 221–229. [Google Scholar] [CrossRef] [Green Version]
  113. Hasan, H.; Shafri, H.Z.M.; Habshi, M. A Comparison between Support Vector Machine (SVM) and Convolutional Neural Network (CNN) Models for Hyperspectral Image Classification. In Proceedings of the IOP Conference Series: Earth and Environmental Science, Kuala Lumpur, Malaysia, 25 November 2019; Institute of Physics Publishing: Bristol, UK, 2019; Volume 357. [Google Scholar]
  114. Kabisch, N.; Haase, D. Green Justice or Just Green? Provision of Urban Green Spaces in Berlin, Germany. Landsc. Urban Plan. 2014, 122, 129–139. [Google Scholar] [CrossRef]
  115. Rigolon, A.; Browning, M.; Lee, K.; Shin, S. Access to Urban Green Space in Cities of the Global South: A Systematic Literature Review. Urban Sci. 2018, 2, 67. [Google Scholar] [CrossRef] [Green Version]
  116. Larson, L.R.; Jennings, V.; Cloutier, S.A. Public Parks and Wellbeing in Urban Areas of the United States. PLoS ONE 2016, 11, e0153211. [Google Scholar] [CrossRef]
  117. Yigitcanlar, T.; Kamruzzaman, M.; Teimouri, R.; Degirmenci, K.; Alanjagh, F. Association between Park Visits and Mental Health in a Developing Country Context: The Case of Tabriz, Iran. Landsc. Urban Plan. 2020, 199, 103805. [Google Scholar] [CrossRef]
  118. Lee, A.C.K.; Maheswaran, R. The Health Benefits of Urban Green Spaces: A Review of the Evidence. J. Public Health 2011, 33, 212–222. [Google Scholar] [CrossRef] [PubMed] [Green Version]
Sustainability 15 00146 g001 550
Figure 1. Spatial distribution map of the urban parks in Shanghai, China included in this study.
Figure 1. Spatial distribution map of the urban parks in Shanghai, China included in this study.
Sustainability 15 00146 g001
Sustainability 15 00146 g002 550
Figure 2. Workflow of sentiment analysis.
Figure 2. Workflow of sentiment analysis.
Sustainability 15 00146 g002
Sustainability 15 00146 g003 550
Figure 3. Conceptual structural model. The number “1” on the arrow refers to a fixed factor loading of 1 on the measurement metric.
Figure 3. Conceptual structural model. The number “1” on the arrow refers to a fixed factor loading of 1 on the measurement metric.
Sustainability 15 00146 g003
Sustainability 15 00146 g004 550
Figure 4. Sentiment analysis inside and outside parks (total: 60 K users).
Figure 4. Sentiment analysis inside and outside parks (total: 60 K users).
Sustainability 15 00146 g004
Sustainability 15 00146 g005 550
Figure 5. Standardized coefficients of the structural equation model.
Figure 5. Standardized coefficients of the structural equation model.
Sustainability 15 00146 g005
Sustainability 15 00146 g006 550
Figure 6. Standardized coefficients of the structural equation model in groups of ‘resident’ and ‘non-resident’. Coefficients (cluster 95% CI) of the SEM.
Figure 6. Standardized coefficients of the structural equation model in groups of ‘resident’ and ‘non-resident’. Coefficients (cluster 95% CI) of the SEM.
Sustainability 15 00146 g006
Sustainability 15 00146 g007 550
Figure 7. Standardized coefficients of the structural equation model in groups of ‘female’ and ‘male’. Coefficients (cluster 95% CI) of the SEM.
Figure 7. Standardized coefficients of the structural equation model in groups of ‘female’ and ‘male’. Coefficients (cluster 95% CI) of the SEM.
Sustainability 15 00146 g007
Table
Table 1. Mediating variables and the categories of data.
Table 1. Mediating variables and the categories of data.
ClassesNumber of People%
≥5 times/month371.00
2–4 times/month3699.92
0–1 times/month331289.08
Total3718100
Table
Table 2. Description of variables measuring urban park attributes.
Table 2. Description of variables measuring urban park attributes.
Latent VariablesObserved VariablesDescription
AccessibilityWalking When the travel mode is walking, biking or driving, the accessibility of a park is based on the Gaussian two-step floating catchment area method (G2SFCA).
Biking
Driving
NaturalnessGreenness coverageThe area of greenness (tree and lawn) is divided by the area of the park.
Water coverageThe area of water body divided by the area of the park.
Table
Table 3. Results of reliability and validity testing.
Table 3. Results of reliability and validity testing.
Latent VariablesObserved VariablesMeanSDCCACRAVE
AccessibilityWalking 0.5170.8870.7250.8730.698
Biking1.5812.057
Driving1.6981.596
Table
Table 4. Standardized effects between factors and expressed happiness.
Table 4. Standardized effects between factors and expressed happiness.
AccessibilityGreenness CoverageAgeWater CoveragePark Visits
TEDEIETEDEIETEDEIETEDEIETEDEIE
Sentiment0.1110.0550.0550.0730.0370.0370.6990.3490.3490.0200.0110.0090.6110.6110.000
Park visits0.0910.0910.0000.0600.0600.0000.5720.5720.0000.0150.0150.0000.0000.0000.000
Note: TE = Total Effect; DE = Direct Effect; IE = Indirect Effect.
Table
Table 5. Results of the tested hypotheses.
Table 5. Results of the tested hypotheses.
HypothesisCorrelationCoefficientResults
H1The greater the accessibility of the park, the happier people who have access to it.0.111Supported
H2The greater the natural degree of the park, the happier people who have access to it.0.093Supported
H3The more often people visit the park, the happier people who have access to it.0.611Supported
H4Disparities between different groups (residents and non-residents, female and male)Existed Supported
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Lai, S.; Deal, B. Parks, Green Space, and Happiness: A Spatially Specific Sentiment Analysis Using Microblogs in Shanghai, China. Sustainability 2023, 15, 146. https://doi.org/10.3390/su15010146

AMA Style

Lai S, Deal B. Parks, Green Space, and Happiness: A Spatially Specific Sentiment Analysis Using Microblogs in Shanghai, China. Sustainability. 2023; 15(1):146. https://doi.org/10.3390/su15010146

Chicago/Turabian Style

Lai, Siqi, and Brian Deal. 2023. "Parks, Green Space, and Happiness: A Spatially Specific Sentiment Analysis Using Microblogs in Shanghai, China" Sustainability 15, no. 1: 146. https://doi.org/10.3390/su15010146

APA Style

Lai, S., & Deal, B. (2023). Parks, Green Space, and Happiness: A Spatially Specific Sentiment Analysis Using Microblogs in Shanghai, China. Sustainability, 15(1), 146. https://doi.org/10.3390/su15010146

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop