Next Article in Journal
Design and Evaluation of Historically and Culturally Integrated Metro Spaces: A Case Study of Xi’an Metro Stations
Previous Article in Journal
Synergistic Role of Recycled Concrete Aggregates and Hybrid Steel Fibers in Roller-Compacted Concrete Pavements: A Multi-Criteria Assessment for Eco-Efficiency Optimization
Previous Article in Special Issue
Exploring the Application of Smart City Concepts in New Town Development: A Case Study of Zhongyang Road, Hsinchu City, Taiwan
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Spatial Optimization of Primary School Campuses from the Perspective of Children’s Emotional Behavior: A Deep Learning and Machine Learning Approach

1
School of Architecture and Planning, Hunan University, Changsha 430071, China
2
School of Architecture and Urban Planning, Huazhong University of Science and Technology, Wuhan 430074, China
*
Author to whom correspondence should be addressed.
Buildings 2025, 15(23), 4281; https://doi.org/10.3390/buildings15234281
Submission received: 3 September 2025 / Revised: 25 October 2025 / Accepted: 5 November 2025 / Published: 26 November 2025
(This article belongs to the Special Issue Research on Health, Wellbeing and Urban Design)

Abstract

From the perspective of children’s emotional behavior, this study constructs a multidimensional indicator framework—“spatial elements-spatial typologies-spatial color-emotion and behavior.” Integrating behavior mapping, we employ deep- and machine-learning models to quantify the pathways through which primary-school campus spaces shape children’s emotional and behavioral responses. The results indicate that: (1) individual external spatial elements exert a more pronounced influence on children’s emotions; (2) different spatial typologies show marked disparities in emotional activation, characterized by polarization and clustering at low levels, revealing common shortcomings in current campus construction; (3) the emotional effects of spatial color diverge systematically by gender and age, with differentiation intensifying as age increases; and (4) overlay analyses of behavior maps corroborate associations between external natural spaces, key internal functional zones, and children’s behavioral patterns and affective responses. Building on these findings, the study proposes targeted optimization strategies oriented toward children’s emotional experience and behavioral development, providing data-driven support for the affective design of primary-school campuses.

1. Introduction

In recent years, the popularization and expansion of primary education worldwide have led to a significant increase in school enrollments. This trend is particularly evident across many countries and regions. As key settings for children’s learning and development, the spatial design of primary school campuses, and its research significance, has become increasingly prominent. Although growth rates vary across countries and regions, the optimization and renovation of primary school campus spaces has become a widely discussed concern.
Over the past five years, China’s primary school enrollment has increased by 6.3%. Among 32 surveyed cities, 28 exceeded the national growth rate, and 18 recorded an increase of 20% or more. Changsha led the country in 2022, recording a growth rate of 41.9%. These figures underscore the research importance and practical value of optimizing primary school campus spaces.
Traditional primary school campus planning has emphasized spatial layout and functional programming. However, it often neglects children’s subjective emotional experiences and behavioral characteristics, leaving substantial room for evidence-based campus spatial improvement. In recent years, integrating deep land machine learning for data analysis and pattern recognition has offered new perspectives and methods to investigate the relationships among children’s emotions, behaviors, and the objective physical environment.
Existing research indicates that spatial elements [1], spatial types [2], and spatial colors [3] exert varying degrees of influence on children’s emotional states [4]. Furthermore, the effects of objective physical spaces vary systematically across gender and age groups [5]. These findings suggest that campus space planning should better account for the diversity and complexity of children’s emotional experiences [6].

1.1. Campus and Students’ Mental Health

The campus, which serves as a crucial entry point for examining the physical environment [7], shapes learning efficiency and mental health through perceptual, cognitive, and affective pathways.
The spatial elements of a campus directly influence students’ psychological perception. Studies based on psychometric scales suggest that the visual characteristics of campus spaces significantly affects individuals’ physical health, emotional recovery, and spatial experience [8]. Additionally, research employing photo-questionnaires and affective-evaluation methods has examined the impact of spatial elements—such as plants, buildings, roads, and water bodies—on landscape experience [9]. Findings indicate that natural elements, particularly plants, play the most significant role in enhancing spatial pleasure [10].
Attention restoration capacity is equally critical to students’ learning efficiency and mental health. Campus spaces with high restorative potential typically feature openness and a balanced integration of natural and artificial elements [11]. Moreover, spaces with well-developed pedestrian systems, high visibility, and well-organized distributions of natural and artificial elements enhance attention restoration [12,13]. Research on restorative properties in primary school campuses has emphasized the importance of multisensory experiences [14], leading to design-optimization principles centered on “naturalness,” “multisensory engagement,” and “future-oriented design” [15]. These principles align with the PAD affective model (Pleasure-Arousal-Dominance) adopted in this study: Pleasure denotes emotional valence, Arousal reflects activation, and Dominance indicates a perceived quantitative entry point for evaluating a campus’s capacity to support students’ mental health and are consistent with the three design principles for primary school environments.
Specific spatial elements play a crucial role in enhancing positive emotions and behaviors, thereby supporting perception and cognitive functions. Previous studies have used environmental variables—such as the Green View Index (GVI), Sound Environment Index (SEI), and Visual Complexity Index (SVI)—in multiple regression models to show that certain spatial configurations significantly improve emotional states [16]. Findings indicate that walking spaces with high GVI and SEI effectively stimulate positive emotions [17], whereas spaces with high SVI and Attention Load Index (AFI) substantially increase negative emotional experiences [18]. These results further support the cognitive dimension of the attention-load theory in campus environments [19]. Learning, leisure, and activity zones—as well as interior versus exterior spaces—differ markedly in cognitive load and social cues. Learning zones typically require configurations that support high levels of sustained focus, whereas activity zones prioritize dynamic settings that foster high-arousal social interaction. Accordingly, it is necessary to examine, at the level of spatial typologies, the distribution of affect across scenarios and the potential for polarization in emotional responses. This study simplified the relevant expressions of the SAM scale and conducted practice trials and comprehension checks.
Mobile-sensing surveys introduce a novel data collection approach for analyzing campus space utilization. By leveraging activity trajectory analysis [20], time window assessments [21], and emotion mapping [22], researchers can identify students’ spatial usage patterns and formulate optimization strategies. This approach enhances the quantitative analysis of campus space usage [23], broadening the scope and accuracy of data collection. This advancement supports robust associations between objective spatial environments and subjective emotional responses, providing a scientific basis for campus planning, design, and optimization strategies.

1.2. Campus Basic Theory

Emotional attachment theory, attentional load theory, and environmental psychology collectively the theoretical framework for this research. Emotional attachment theory emphasizes the emotional bonds that children establish with their environment, shaping how they perceive and interact in campus settings. Attentional load theory articulates links between space design and children’s behavioral health and examines how environmental complexity affects cognition and behavior from psychological and neuroscientific perspectives [24,25]. Environmental psychology investigates the interplay between individuals and their physical and social surroundings, focusing on how context shapes behavior, emotions, and cognition [26]. Given ongoing cognitive and neurological development, children are particularly sensitive to environmental stimuli. Overly complex conditions—such as crowding, harsh lighting, and loud noise—substantially increase cognitive and attentional load [27], leading to reduced attention spans and, at times, emotional disturbances [12].
Place attachment theory underscores the profound influence of place on behavior and emotional development [28]. Originally articulated by Williams et al. in 1989, the theory integrates concepts from psychology and geography, and posits that people form emotional and cognitive bonds with specific places [29]. It comprises two key components: place dependence and place identity [30]. Place dependence concerns functional reliance on a space to meet particular needs [31], whereas place identity refers to an emotional sense of belonging and identification with place [32].
Place attachment theory offers a critical perspective for examining children’s emotions and behaviors [33], as their emotional ties to school grounds directly shape their behavioral patterns [34]. By analyzing how children establish subjective emotional attachments to schools under the influence of objective physical factors—such as spatial elements, spatial colors, and spatial types [35]—researchers gain a theoretical basis for understanding the mechanisms through which the objective physical environment affects children’s emotional and behavioral responses. Attachment theory, attentional/cognitive load theory, and environmental psychology provide the study’s cognitive framework. In primary-school settings, maintaining a balance among “restoration-activation-control” is the central mandate of a mentally healthy campus. To translate children’s affect-behavior health into verifiable evidence, we (i) employ AI-based methods to obtain repeatable measurements of the objective physical environment, and (ii) use the PAD affect model together with children’s behavior maps to capture both subjective emotion and actual patterns of use. Linked through machine-learning analyses, these data identify dominant factors, heterogeneous effects, and potential thresholds, from which we propose actionable spatial optimization and color strategies.

1.3. Research on School Space in Primary Schools

The shapes, combinations, and proportions of design elements exert varying influences on children’s emotions and behaviors [36]. Survey evidence indicates that campus greenery is positively correlated with children’s well-being and perceived quality of life [37,38,39]. Moreover, design features such as desks, chairs, decorative fixtures, doors, windows, suspended ceilings, walls, and floors affect children’s stress and learning behaviors, contingent on the selected colors, materials, and coverage areas [40,41,42,43,44,45,46].
Studies of Canadian and European university campuses suggest that student well-being and satisfaction are associated with campus location, public art, and the availability of essential amenities [47,48]. Another study reported that adding cartoon decorations to indoor settings increased the duration and frequency of children’s attention and reduced anxiety about medical visits [49,50]. These findings imply that appropriate interior decorations [51] can help create a comfortable indoor environment and stimulate positive emotions in children.
Given the strong link between physical activity and overall health [52,53], sports field design plays a pivotal role in children’s development by providing spaces necessary for growth [54]. Informal outdoor areas enable creative activity [55] and facilitate physical and cognitive development through movement and play in natural settings [56,57].
Leveraging semantic segmentation in deep learning [58], this study categorizes campus spatial design elements [59] into two primary groups: indoor and outdoor [60]. The indoor category comprises nine sub-elements: tables and chairs, cabinets and furniture, floor coverings, ceilings, enclosing walls, doors and windows, facilities and decorations, interior landscaping, and building elevations. The outdoor category comprises eight sub-elements: floor coverings, road paths, playgrounds, building elevations, enclosing constructions, facility guiding systems, green landscaping, and sky view [61].
Color [62,63] and material [64] can significantly affect individuals’ mental state and perception [65]. Studies on color perception suggest that red or orange interiors typically evoke stronger positive affect than purple or blue interiors [66]. Lighter colors are perceived as friendlier, brighter, and more uplifting, making everyday activities feel easier and more enjoyable [67]. Similar color schemes may yield comparable benefits for children, enhancing positive moods and stimulating constructive cognitive activity. Experimental evidence indicates that altering interior materials has notable effects; for example, adding wooden components to learning spaces reduces fatigue, increases pleasure, and improves the capacity to refocus [68]. Thus, introducing additional wooden materials may improve learning environments for school students.
This study follows established indicators for color in architectural spaces [69] and their corresponding calculation methods [70], just as the indicators shown (Table 1).
To quantitatively assess how children’s emotions are influenced by different spatial forms of campus environments [71], this study introduces the Pleasure-Arousal-Dominance (PAD) emotion model [72] to analyze emotional states along multiple dimensions. Drawing on Osgood’s research [73,74], emotional experiences are evaluated along three key dimensions: evaluation, potency, and activity [75]. building on this framework, Mehrabian and Russell proposed a three-dimensional model of emotional states [76]. These dimensions correspond to experiential needs in campus spaces: restoration, activation, and control [77,78,79].
By quantifying children’s emotional perceptions of campus space [80] through the PAD model, this study systematically evaluates the effects of primary school campus environments on children’s emotions [81,82]. Building on previous research, it categorizes children’s emotions into Arousal, Dominance, and Pleasure [83,84,85]. The study quantifies children’s emotional appraisal of different spatial scenarios using PAD scores and cross-validates these with the spatiotemporal usage patterns derived from behavior maps. This approach enables an examination of how primary-school campus spaces shape children’s emotions and behaviors, while testing for heterogeneity by gender and age. Continuing the three-stage framework of “emotion recognition-emotion processing-emotion expression,” emotions are represented along three dimensions—Pleasure, Arousal, and Dominance.
Children’s emotional needs vary across spatial types [86], shaping distinct patterns of behavior [87]. Campus public spaces, such as green spaces, corridors, and classroom areas, serve different emotional requirements and thus affect children’s behavior in diverse ways [88,89,90]. Therefore, differentiating these spaces is essential for precise quantitative analysis [91,92].
To systematically analyze how spatial environments influence on children’s emotional responses, this study categorizes campus spaces into two primary categories: internal and external.
Internal space comprises three functional zones:
(i)
Study areas: library, study classroom, art classroom, multifunctional classroom, and lecture hall.
(ii)
Living and leisure areas: restroom, cafeteria, atrium, auxiliary area.
(iii)
Activity areas: sports field, gallery space, planting garden, stairwell, and shared activity space.
(iv)
External spaces include building facades (inner courtyards, activity areas, athletic areas, elevated areas, outdoor character areas), enclosure components, utilities, exercise paths, and temporary features [93].
The innovations fall into two aspects: perspective and methodology.
(i)
Perspective: This study adopts a user-centered approach, emphasizing children’s emotional development and behavioral performance within the school environment. This perspective addresses the paucity of systematic research on elementary school campuses that explicitly considers children’s experiences [94]. By integrating spatial elements, spatial types, and color characteristics from the objective physical environment with students’ subjective emotional perceptions and behavioral patterns, the study constructs a comprehensive influence mechanism model [95]. This approach overcomes the limitations of prior studies that examined these factors in isolation.
(ii)
Methodology: The study employs deep learning and machine learning techniques [96,97,98,99,100,101,102,103] to establish relational models linking emotion and behavior. This approach provides empirical support and practical guidance for campus space design, enhancing the accuracy and applicability of the findings.

2. Methods

2.1. Date

This study examines Yuying Huizhan Primary School, Shazitang Primary School, and Bai Ruo Pu Guangming Primary School as case schools. These three public elementary schools in Changsha, Hunan Province, have strong educational resources and follow widely adopted, representative educational practices. Geographically, they are located in distinct areas of Changsha—Yuhua District, Changsha County, and Wangcheng District—thereby reflecting the city’s diversity in elementary education.
The study ran from 1 September 2024 to 20 January 2025, using both online and offline methods. A total of 390 students participated: 194 boys (49.74%) and 196 girls (50.26%). To ensure stable in emotional responses and reliable data, participants had to have lived in Changsha for ≥6 months and attended school for >3 months, ensuring familiarity with the school environment.
The research comprised three components: Affective questionnaire, Virtual simulation experiment, and behavior map monitoring. In total, 3850 comparative data sets were collected for the three affective indicators—pleasure, arousal, and dominance—yielding 11,550 pairwise comparisons. The questionnaire assessed how school spaces affect children’s emotions and behaviors by comparing and contrasting choices across the PAD dimensions. The survey used simplified language and visual contrast diagrams to enhance comprehension. Data were collected online and offline with assistance from teachers, principals, and volunteers.

2.2. Framework

This study investigates the mechanisms linking children’s emotions and behaviors with the objective built environment, integrating two dimensions: the objective built environment, which encompasses spatial elements, spatial types, and spatial colors, and subjective behavioral experiences, which pertain to children’s emotional responses and behavioral patterns. The research systematically examines three key aspects: how the objective built environment affects children’s on-campus emotions; correlations among spatial elements in the built environment; and superimposed analysis of built environment elements with children’s behaviors.
The study comprises four stages (Figure 1). First, during data collection, fixed-position indoor and outdoor cameras capture spatial information, and surveillance records plus photographic observations document behavioral data. Second, during data processing, color correction is performed via automatic white balance (AWB) in OpenCV, and deep learning semantic segmentation is used to extract spatial elements and compute spatial and color indices. Third, during quantification, children’s emotional and behavioral data are analyzed: choice-based emotional tendencies are mapped to the PAD dimensions, and quantified by recoding activity type, duration, and frequency in specific spaces through observation. Finally, based on the findings, targeted spatial strategies are developed to optimize spatial types, spatial elements, and spatial colors, and these are superimposed on behavior maps to provide practical design guidance for campus planning.

2.3. Methodology

(1)
Semantic segmentation based on deep learning
The study employs the Segment Anything Model (SAM) to perform semantic segmentation on real-world campus imagery, automatically identifying indoor and outdoor spatial elements. In artificial intelligence, the Segment Anything Model (SAM) is a major segmentation, exhibiting strong adaptability across diverse tasks. Built on Transformer-based architecture, SAM achieves pixel-level segmentation using a large-scale training corpus (Figure 2).
In this study, SAM enables fine-grained segmentation of spatial elements by accurately identifying components such as interior spaces, home decorations, and window assemblies. Its standardized segmentation performance ensures consistency across images, making it well suited to processing large-scale building image datasets.
(2)
Machine Learning-Based Modeling of Influence Mechanisms
Random forest regression was used to model, separately, the effects of spatial elements, typologies, and color on the three PAD dimensions. This approach accommodates nonlinearity and interactions, while providing feature importance and response profiles to identify key factors and threshold ranges. The random forest model demonstrated strong performance and high predictive accuracy in both training and prediction phases. The Random Forest Regression (RFR) is used to learn a predictive mapping that links objective physical environment to children’s subjective emotions in primary school campuses. This approach seeks to uncover relationships among spatial elements, spatial types, and color characteristics and children’s emotional responses.
In this study, spatial elements and color characteristics, extracted via semantic segmentation serve as independent variables, whereas the three PAD dimensions serve as dependent variables to construct a spatial-emotional model for elementary school campuses. To enhance stability, hyperparameter optimization are optimized and k-fold cross-validation is conducted to improve out-of-sample generalization. Results indicate that the RFR achieves strong performance for modeling and predicting of children’s emotional responses.
(3)
Behavioral map observation based on video images
Behavior maps were constructed using video footage and systematic time-interval sampling to identify spatial-use patterns—including time, location, participants, frequency, and duration—which were cross-validated against the PAD results.
This study analyzed children’s behaviors during the first two weeks of November 2024 using video recordings to conduct behavioral map observations. Recording sessions were scheduled during stable, sunny periods to minimize temperature fluctuations. Observations spanned seven consecutive days and focused on elementary school recess periods; typical spatial locations were selected for analysis. Behaviors lasting ≥3 s were recorded [12], and repeated behaviors separated by ≥30 s were classified as new to ensure methodological rigor.
To enhance accuracy, the study used timed sampling with 10-min sessions. Manual recording and video playback analysis were used to compile detailed statistics on behavior types and frequencies [11]. To reduce stochastic variation, each space was observed across 10 consecutive intervals, and results from multiple recordings were averaged to improve stability and representativeness.
Observers were assigned to each site to record children’s outdoor activities, including participant gender, activity duration, and stay points. Additionally, fixed cameras were installed at observation sites to assist behavioral recording and ensure comprehensive data capture.
The final mean count of recorded behaviors was 884, with 412 behaviors occurring in external spaces and 472 in internal spaces. These data (Table 2) reveal the spatial distributions of children’s behaviors and provide empirical support for understanding how elementary school campus space influence students’ emotions and behaviors.

3. Results

3.1. Influence of Spatial Elements on Children’s Emotions

Children’s emotional responses vary markedly across indoor and outdoor spatial elements (Figure 3).
Arousal is strengthened most by targeted spatial interventions and broader sightlines; indoors, cabinets and floor finishes exert strong effects, with ceilings and seating also contributing, underscoring the cumulative role of interior detailing. By contrast, outdoor elements such as sky and greenery show weaker arousal, likely due to their indirect engagement and field-of-view/distance constraints.
Dominance increases with definition and visibility cues: seating ranks first (24.10%), and doors/windows also contribute, indicating that furniture configuration and openings enhance children’s perceived control and spatial belonging, whereas wayfinding systems play a secondary role.
Pleasantness is driven primarily by indoor elements; ceilings have the largest effect (21.20%), followed by enclosing walls and seating, suggesting that large-scale interior components more strongly shape a warm, pleasant atmosphere. Outdoor paths and sports fields display diffuse, non-dominant effects.
It is evident that elementary school campus building space design should adhere to three key principles:
(i)
Prioritize high-impact elements: Treat components that influence multiple emotional dimensions as primary design targets.
(ii)
Layer functions of spatial elements: Because elements contribute differently to emotional responses, define their functional hierarchy and coordinate design according to their roles within each dimension.
(iii)
Integrated across factors: No single element can meet children’s emotional needs; therefore, employ multi-factor, synergistic strategies to optimize the overall emotional experience.

3.2. Influence of Space Type on Children’s Emotions

Children’s emotions vary by campus spatial type (Table 3).
Interior study spaces show low arousal and pleasure with polarized dominance, indicating monotony and insufficient stimuli for engagement and emotional diversity.
Leisure and living areas also skew toward lower emotional levels, reflecting simple layouts that limit expressive behavior and social interaction, thereby weakening emotional participation.
Activity areas exhibit a bimodal distribution (both high and low), but with notably higher pleasantness, suggesting designs that prioritize interaction and flexibility, which support emotional well-being despite some residual spatial constraints.
The study’s findings focus on the PAD arousal dimension rather than overall affect: data show that external natural elements (sky view, distant greenery) are associated with lower arousal but not lower pleasure, which aligns with restorative theories in environmental psychology—these theories hold that nature tends to calm rather than excite people. In outdoor spaces, arousal and pleasure remain low while dominance is strongly polarized, implying that current configurations do not effectively elicit participation or interest.
The dominance polarization further signals limited autonomy and control, likely arising from irrational layouts, inappropriate color use, and excessive fixed or overly enclosed elements that reduce flexibility.
Overall, external spatial elements and color schemes in the surveyed Changsha schools are insufficiently stimulating, underscoring the need for targeted improvements to align campus design with children’s emotional and psychological needs.

3.3. Influence of Spatial Color on Children’s Emotions

(1)
Gender-Based Characteristics of Children’s Emotional Responses
Color indicators affect the three emotional dimensions differently for girls and boys (Figure 4).
For arousal, girls are more sensitive to saturation (C8) and hue contrast (C4), whereas boys respond more to luminance/brightness contrast (C6); color warmth–coolness exerts little influence on boys’ arousal.
On the arousal dimension, for girls, hue contrast (C4), hue index (C7), and luminance (C9) show stronger effects, indicating that primary hue choices and overall luminance shape perceived control. Boys show a marked preference for luminance (C9) and a more even response across indices, suggesting broader tolerance to color variations when establishing control.
For pleasantness, girls show heightened sensitivity to hue (C7), luminance contrast (C6), and hue contrast (C4), implying that soft, layered color schemes most effectively enhance pleasantness. Boys’ pleasantness is influenced by C4, C6, C5 (saturation contrast), and, to a lesser extent, C7 and C8, indicating benefits from vivid, well-defined tonal atmospheres.
Warm–cool tendency contributes to ambience but has weak, non-specific effects on discrete emotional expressions for both genders.
Tailor interior color strategies by gendered sensitivity profiles—emphasizing hue layering and saturation for girls, and luminance/contrast and vivid tonal structure for boys—while de-emphasizing warmth–coolness as a primary lever for specific emotions.
(2)
Children’s Emotional Characteristics by Age
Figure 5 contrasts color–emotion associations across ages 7–8, 9–10, and 11–12.
Children aged 7–8 show immediate, high-arousal responses driven by color expressiveness (B3) and brightness index (C9), amplified by dominant-color count (C1) and ratio (C2); this rapid engagement likely elevates attentional load and may impede restoration. Dominance is comparatively stable and shaped by visual impact (B2), while pleasantness tracks hue, saturation, and brightness contrasts (C4–C6).
Among 9–10-year-olds, responses diversify: arousal becomes more sensitive to Brightness Index (C9) and saturation (C8), whereas the effects of C1, C2, and C6 weaken relative to ages 7–8; pleasantness increasingly reflects atmospheric qualities captured by B2 and B3 and benefits from richer combinations rather than greater quantity.
By ages 11–12, patterns appear more mature: arousal is driven by primary color ratio (C2), hue contrast (C4), and luminance contrast (C6), and the influence of color richness (B1) strengthens, indicating more sophisticated appraisal that balances aesthetics with comfort.
Overall, with age the role of color shifts from provoking immediate stimulation to supporting multidimensional appraisal, regulation, and self-adaptation at the level of spatial atmosphere and function.

3.4. Correlation Between Spatial Elements and Spatial Color

A Pearson correlation analysis was conducted to examine the relationship between internal and external spatial elements and color indicators in primary school campuses (Table 4). This section is only used as an exploratory descriptive screening for the study’s variable selection, and the study’s claims do not rely on p-values. This study examines nine internal spatial elements, including tables and chairs, cabinet furniture, floor pavement, ceiling, enclosed wall facade, door and window construction, facilities and decoration, indoor greening, and building facade, as well as eight external spatial elements, including ground pavement, road path, sports field, building facade, enclosure construction, facility guide system, green landscape, and sky vision, in relation to ten color indicators. The findings indicate that spatial color influences children’s visual perception and directly shapes their emotional responses.
Hierarchical adaptation of color complexity to children’s emotional needs is evident in how specific spatial elements interact with color attributes. Tables and chairs show significant negative correlations with hue contrast C4 and brightness contrast C6, suggesting that low-contrast color schemes attenuate visual stimulation, thereby creating a calmer and more stable atmosphere. Cabinet furniture shows positive correlations with the saturation index C8 and brightness index C9, suggesting that its color design enhances visual richness and improves color differentiation within the space.
The identification of spatial elements and their role in defining emotional needs can be interpreted through place attachment theory, which emphasizes the connection between spatial recognizability and children’s emotional attachment. Doors and windows show significant negative correlations with color quantity C1, color harmony type C3, and brightness contrast C6, and a positive correlation with color warmth C10. These findings suggest that warmer color schemes in doors and windows contribute to a more inviting and comfortable spatial atmosphere. However, current color applications for doors and windows remain monotonous and insufficiently consider how different colors influence children’s spatial perception. Reduced color diversity weakens visual guidance by doors and windows, whereas color contrast diminishes spatial layering, making the environment appear dull and less engaging. Balancing minimalist design, spatial recognizability, and functional definition remains essential for optimizing current school spaces.
The systematic coupling of color attributes with emotional dimensions underscores the importance of brightness and saturation in children’s emotional responses. An increase in the brightness index C9 was associated with significantly higher pleasantness, reinforcing a strong association between brightness and emotional experience. According to environmental psychology and behavioral research, bright hues can enhance positive emotions, whereas excessive brightness may include inattention and visual fatigue in some children. Therefore, efforts to improve spatial pleasantness should also mitigate potential adverse effects of excessive brightness to support emotional stability. Future research should further investigate strategies for balancing color intensity to optimize emotional engagement and environmental comfort for children.

3.5. Overlay Analysis of Children’s Behavioral Maps

(1)
Campus external space
Using aligned overlays of behavior maps, activity modes/frequencies, and the proportional composition of spatial elements for each setting (Table 5), we identify latent associations among children’s activity types, spatial behavior patterns, spatial elements, and color features in outdoor campus areas.
Sports zones and public activity spaces show high rates of sport and group interaction, where sky view, greenery, and ground paving predominate. Overhead and outdoor feature areas are characterized by observation, strolling, and small-group socializing, with architectural façades and sports grounds forming the principal elements. These patterns indicate that functional orientation and environmental quality directly shape behavioral choices and participation: open sightlines and sports facilities promote physical and social activity, whereas richer landscapes and sheltered spaces foster close interaction and individualized use.
Across settings, seating/furniture, floor paving, and enclosing components significantly influence both activity type and behavioral intensity. Inner courtyards with strong enclosure encourage focused play, quiet rest, and brief retreats; in sports areas and along sports paths, pavement quality correlates with the occurrence and frequency of sport. Conversely, in overhead zones and some public spaces, insufficient public facilities and decorative elements constrain opportunities for outdoor learning and exploration.
In the behavior map, children’s behavior symbols are represented as follows:
Sitting and resting: Buildings 15 04281 i004
Observing and waiting: Buildings 15 04281 i005
Reading alone: Buildings 15 04281 i006
Walking with companions: Buildings 15 04281 i007
Playing games: Buildings 15 04281 i008
Spectating (onlookers): Buildings 15 04281 i009
Discussing with teachers: Buildings 15 04281 i010
Reading with peers: Buildings 15 04281 i011
Sports activities: Buildings 15 04281 i012
(2)
Campus internal learning space
By superimposing behavior maps from libraries, learning classrooms, and multifunctional rooms, we find that children’s behavioral patterns in internal learning spaces are tightly linked to spatial elements and color configurations (Table 6).
Libraries and multifunctional rooms exhibit high shares of quiet behaviors—reading, discussion in low voices, resting, quiet reflection, and observation—dispersed across discrete zones. This pattern suggests that comfortable, low-noise layouts with clear visual foci foster immersive learning; these spaces typically feature high-quality seating/furniture and acoustic tile ceilings that enhance both function and aesthetics. In learning classrooms, discussion and interactive behaviors concentrate in central areas or around group seating, indicating that flexible desk–chair arrangements support teacher–student communication and collaborative learning.
Linking spatial elements to emotion, floor finishes and ceiling systems coincide with hotspots for reading and discussion, implying that soft palettes or localized accents can improve comfort and concentration. Moreover, moderate transparency or strategic partitioning in walls and windows supports differentiated emotional needs across learning activities. Finally, peer walking and game breaks are infrequent within learning spaces, indicating a lack of designated relaxation zones and a corresponding gap in meeting diverse emotional needs during breaks.
(3)
Living and leisure space inside campus
A superposition analysis of behavior maps for living and leisure areas (Table 7) shows three high-frequency patterns: short stays, social interaction, and walking/observation.
Toilets and adjacent zones function primarily as transit spaces with minimal dwelling or conversation, indicating fulfillment of physiological needs but limited psychological comfort; color palettes are monotonous, and although privacy, ventilation, and lighting meet basic requirements, emotional support is underprovided.
The canteen, dominated by enclosing façades and floor finishes, supports seated resting and peer conversation, effectively facilitating group dining and socialization but offering insufficient provision for quiet individual activities, small games, or reading.
By contrast, the atrium, with its open layout and high permeability, elicits exploratory and interactive behaviors (walking with peers, informal play), reflecting strong environmental stimulation and elevated movement-based and social activity.
Across living and leisure settings, watching, solitary reading, and other individualized uses remain scarce, suggesting a need for designated quiet/independent zones and richer, child-oriented color and detail strategies to broaden emotional engagement.
(4)
Internal Activity Spaces
Overlaying behavior maps for five indoor activity settings—plantation, sports field, corridor, stairwell, and shared activity space (Table 8)—shows that these environments stimulate curiosity and social interaction but also demand strong cues for safety and spatial legibility.
The sports field yields the highest rates of leisure sport and peer play, indicating that open, dynamic layouts promote fitness and cooperation. The plantation encourages observation and hands-on planting, but affords limited in-depth teacher–student interaction, suggesting scope for instructional features within nature-based areas. Corridors and stairwells, as semi-open and vertical circulation zones, support strolling, observing, and brief play; elongated corridors with appropriate color, floor markings, and wayfinding enhance interest and exploration, whereas poor permeability and guidance shorten dwell time. Given their role in movement, these zones also require improved visibility, elimination of blind spots, and other safety measures. The shared activity space accommodates discussion, play, solitary reading, and relaxation/exercise, demonstrating the value of flexible layouts and adaptable furniture; however, high traffic necessitates carefully planned circulation routes and clearer spatial markers. Monotonous color schemes and weak color-based zoning reduce functional clarity and can trigger behavioral overlap, limiting exploratory and social engagement.

4. Discussion

4.1. Recommendations from Spatial Elements

Different spatial elements play multidimensional roles in children’s emotional experiences, influencing the three key emotional dimensions in distinct ways. Accordingly, we advocate enhancing the target affective dimensions while controlling extraneous load, thereby avoiding the dual risks of “overstimulation” and “understimulation.”
(1)
Affective calibration and load management of key spatial elements.
In early design, priority should be given to spatial elements that effectively elevate emotional arousal, consistent with attentional load theory and environmental psycho-behavioral theory. Emphasis should be placed on cabinetry and floor finishes, using color contrasts and salient visual focal points to help children engage quickly without undue attentional load.
To foster dominance and belonging, defining elements such as tables, chairs, doors, and windows should be prioritized. These features reinforce spatial legibility and security, mitigating cognitive fatigue associated with environmental disorganization. Well-structured layouts enhance children’s emotional stability and support an engaging learning environment.
(2)
Functional stratification and coordinated configuration of spatial elements.
Different spatial elements play distinct roles in children’s emotional experiences and cognitive load management. In overall design, indoor and outdoor components should be layered and integrated according to their functions. Indoors, ceilings, walls, and furniture can form an emotionally stimulating subsystem, whereas outdoor spaces should function as an auxiliary system characterized by openness and landscape coherence. Cohesive integration of color, material, and form is essential to harmonize these subsystems and to prevent unintended overload of children’s cognitive resources from overstimulating individual elements. The goal is to establish a dynamic equilibrium between emotional engagement and cognitive stability.
Additionally, modular design can accommodate children’s diverse needs while avoiding attentional instability and emotional exhaustion caused by excessive layering of spatial elements. This approach keeps the built environment adaptive and stimulating without overwhelming children’s cognitive capacities. Introduce micro-restoration units at locations prone to sustained cognitive load to mitigate fatigue accumulation. Use modular construction to enable rapid reconfiguration across grades and time periods, providing differentiated affective support without increasing environmental complexity.
(3)
Adopting a Comprehensive Strategy of Multi-Scale and Multi-Scene Design
A balanced spatial strategy should serve both focused learning and restorative areas, enabling seamless transitions between concentration and recovery. Consistent with attentional load theory and environmental psychology, classrooms should incorporate soft color palettes, localized accents, and designated microl-rest areas. These features help maintain attention while enabling rapid state shifts, thereby reducing cognitive fatigue from prolonged high-load tasks.
In activity areas, design should accommodate locomotor play and social interactions. Clear circulation zoning, salient wayfinding cues, and dynamic spatial markers can guide children toward diverse physical and social engagements. This approach helps children maintain optimal attentional load while sustaining appropriate arousal, fostering active participation and well-regulated affect.

4.2. Suggestions from Space Type

According to developmental and behavioral psychology, children’s emotional needs and behavioral patterns across developmental stages are closely linked to the functional organization of their environments. During elementary school years (ages 6 to 12), children accumulate experiences, consolidate self-awareness, and increasingly differentiate attributes and affective cues across campus spaces.
(1)
Learning zones: coordinating sustained focus and controllability
The primary goal of learning zones is to sustain stable concentration alongside a sense of control. From a cognitive-load perspective, extraneous stimuli should be minimized to prevent inefficient use of attentional resources. From environmental psychology, predictable interface rhythms and a consistent cue system reduce uncertainty and anxiety. Accordingly, low-saturation, low-contrast palettes with a controlled number of color categories are recommended to keep visual information “necessary and sufficient,” thereby suppressing extraneous load.
In terms of Dominance, variation in learning zones often stems from differences in children’s control over personal territory. Clarifying individual boundaries and operational cues enhances place control and self-management, strengthening positive attachment. Micro-restoration nodes embedded at corridor ends and corners can interrupt sustained cognitive load through attentional recovery, supporting emotional reset between classes and mitigating exhaustion and attachment erosion associated with prolonged study.
(2)
Leisure and living zones: social activation with low-load pleasure
Leisure zones aim to support peer interaction through low-load pleasurable experiences. Based on cognitive-load theory, compositions should combine focal high-salience points with low-salience backgrounds: localized theme corners and participatory installations provide attraction and organizational structure, while background surfaces remain subdued to avoid over-arousal. Environmental psychology highlights the role of affordances in initiating social behavior: components that are sit-able, reliable, and manipulable offer clear scripts for interaction, lowering barriers to engagement, fostering positive peer experiences, and accumulating place-based emotional attachment.
(3)
Activity zones: order and safety within high arousal
Activity zones are inherently high-arousal settings and therefore require ordering cues to prevent overload and conflict. A triadic system of circulation-zoning-signage should be established: clear pathways link subzones of varying intensity and function, supported by hierarchical functional colors/symbols to reduce aimless traversing and mutual interference. This ensures that high arousal does not devolve into cognitive congestion or negative effects. To address time-specific crowding, incorporate adjustable boundaries—such as soft partitions and mobile elements—as safety buffers and flow-management tools. Predictable control over rhythm and boundaries strengthens children’s sense of dominance and situational safety, enabling high-arousal experiences to coexist with positive attachment.
(4)
Outdoor spaces: enhancing participation and perceived controllability
Outdoor spaces often present higher participation thresholds and fewer controllability cues. Environmental psychology suggests lowering entry barriers through behavioral gradients: along continuous paths, provide nodes for pausing, playing, and observing so that “passing by” naturally transitions into participation. Regarding scale and enclosure, avoid overly hard or closed forms; use visible edges and soft delineation to create legible, safe boundaries without suppressing exploration. This supports a positive experiential pathway that is visible, approachable, and controllable. As cycles of lingering, interaction, and revisiting become established, positive emotional memories consolidate at specific nodes, gradually strengthening children’s attachment to outdoor areas and increasing perceived dominance.
For instance, a transitional zone between study and leisure areas enables gradual emotional shifts; soft color palettes and calming accents facilitate psychological adjustment. Similarly, in activity areas, distinct color boundaries and furniture configurations help maintain functional independence, establishing a well-defined, layered, and complementary spatial hierarchy. Systematic zoning supports children’s emotional stability, cognitive focus, and social engagement, fostering a holistic, enriching campus environment.

4.3. Suggestions from Space Color

Spatial color and spatial elements do not operate independently in shaping children’s emotional and behavioral patterns; instead, they interact across multiple dimensions to form a complementary, tightly coupled system. Therefore, their integration should be addressed systematically in overall design to enhance cognitive and emotional engagement.
(1)
Matching Space Function with Color Characteristics
The match between space functions and color characteristics should reflect each space’ attributes to determine an appropriate color index scheme. To stimulate vitality and encourage social interaction, color contrast can be increased to promote active participation and group engagement. Conversely, in immersive areas, low saturation and soft contrast should be used to minimize visual noise and enhance concentration. Additionally, research in developmental and environmental psychology indicates that zoning should account for age and gender related differences. In lower-grade teaching areas and public corridors, more salient colors can capture attention, whereas in upper-grade spaces, contrast should be moderately reduced and layered palettes introduced to balance functional demands with emotional regulation and aesthetic quality.
(2)
Coupling Spatial Elements with Color Design
Coupling spatial elements with color should prioritize components that strongly affect children’s emotions while aligning with age-specific spatial and color preferences of achieve joint optimization of function and emotion. For instance, tables and chairs influence children’s sense of dominance and emotional stability; therefore, high-lightness moderately saturated colors can create a lively but non-overstimulating atmosphere. Additionally, large-scale elements such as ceilings play a critical role in emotional stimulation. Accordingly, incorporating attentional load theory, spatial emotional demands, and children’s behavioral characteristics can guide color zoning and patterning, ensuring that visual focal points are placed effectively within decorative elements. Effective integration of spatial elements and color indices enhances the practical application of campus design and supports long-term visual adaptation and balance, fostering an environment that sustains children’s emotional and cognitive well-being.

4.4. Suggestions from the Behavior Map

The behavior map offers critical insights into children’s activity patterns and emotional responses across functional areas, enabling targeted spatial optimization to address their diverse physical and psychological needs. Children’s interactions with the built environment are shaped not only by the allocation of spatial elements, color coordination, and functional cohesion, but also by place attachment to the campus. Strengthening this attachment through design can enhance children’s sense of belonging and active engagement, fostering a supportive, enriching school environment.
(1)
Emotive renewal of high-frequency nodes.
In social hubs such as cafeterias and atria, use moderate color temperatures and a restrained palette to reduce extraneous load, and provide sit-able, reliable, and participatory installations with unobtrusive wayfinding to establish clear social affordances. In restrooms, employ bright directional cues and soft décor to lessen avoidance. Meanwhile, organize “pause-and-restart” paths along the main circulation spine: embed “micro-narrative nodes”—including graphic markers, small themed corners, and micro-greening lean-points—to support brief restoration and orientation. These interventions convert repetitive passage into positive memory and place attachment, while concurrently elevating Pleasure and Dominance.
(2)
Enhancing Functional Layers and Decorative Elements in Living and Leisure Spaces
To optimize the functional structure and aesthetic quality of living and leisure areas, fostering an engaging social emotional atmosphere in high-traffic spaces such as canteens and atriums is essential. This can be achieved through playful interventions and strategically placed color accents that encourage peer interaction and social bonding. Additionally, warm lighting schemes and dynamic accents should be integrated into basic functional spaces to provide physiological comfort and emotional security. Consistent with place attachment theory, outdoor emotional engagement should be reinforced via multifunctional activity installations and nature-based experience zones, enabling children to develop positive emotional associations with the campus. Through repeated exposure and interaction, children gradually cultivate deeper attachment and long-term emotional connection to their school surroundings.
(3)
Strengthening Spatial Flexibility and Visual Guidance in Activity Areas
To improve spatial adaptability and wayfinding in activity areas, a flexible, responsive design strategy should be prioritized to support children’s engagement in sports, exploratory play, and group interactions. Legibility and emotional resonance can be enhanced by strategic wayfinding systems, including clear signage, spatial markers, and color-coded elements in corridors, stairwells, and shared activity zones. Adopt a hierarchical signage system—Level 1: Direction, Level 2: Function, and Level 3: Detail—to ensure progressive legibility and mitigate negative affect arising from disorientation and congestion. This multisensory guidance, integrating visual, cognitive, and emotional stimuli, fosters positive place attachment and reinforces the campus long-term educational and developmental role in the children’s holistic growth.

5. Conclusions

The study investigates the mechanisms linking the objective physical environment of elementary school campuses with children’s subjective emotional experiences and behavioral patterns, using deep learning and machine learning methods. We conducted a systematic analysis to examine how spatial elements, spatial types, and spatial colors shape children’s emotional development and behavioral responses. By overlaying findings with behavioral maps, the study examines how school environments meet children’s emotional and behavioral needs, and identifies disparities and opportunities for improvement.
The findings are organized across four levels.
Element level. Large-scale interior elements exert a more direct influence on Pleasure and Arousal, while defining elements such as doors and windows more strongly enhance Dominance. A layered optimization strategy—define-engage-control—is recommended to foreground operable, participatory interfaces and facilities.
Type level. Learning zones tend to show low stimulation and low emotional arousal. Activity and living zones can elicit positive emotions when they are semi-enclosed and allow adjustable intensity. External spaces show clear deficits in participatory potential and perceived control. We recommend refined affective zoning: create focused learning scenes with low saturation, low contrast, and predictable rhythms, and strengthen participation and controllability in external spaces.
Color level. There are gender- and age-related differences. Girls are more sensitive to hue and saturation, while boys are more sensitive to lightness and contrast. With age, preferences shift from high-stimulus palettes to multi-layered, composite needs. A grade and gender responsive scheme is advised, following function first, clear hierarchy, and moderate contrast.
Behavioral integration. Behavior maps indicate that linking external nature with internal transitional spaces, and organizing nodes along the sequence path-pause-re-engage, improves behavioral flexibility and emotional recovery. We recommend implementing this spatial continuity in parallel with the element, type, and color strategies.
This study has several limitations. The sample was drawn from three primary schools in Changsha, and data collection was concentrated in a single stable-weather week in November, which constrains external validity. Future research should conduct longitudinal surveys across multiple cities and climate zones, spanning seasons and times of day; integrate sensor-based data with behavior tracking; and iteratively validate findings through real-world retrofits and post-occupancy evaluation to enhance generalizability and transferability.

Author Contributions

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

Funding

This research was funded by the Natural Science Foundation of Hunan Province of China, grant number 2024JJ3010, and by the Hunan Office of Philosophy and Social Science, grant number 23WTC03, and by the Natural Science Foundation of Hunan Province of China, grant number 2023JJ40144.

Data Availability Statement

The data presented in this study are available on request from the corresponding author due to the complexity of the data.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Pallasmaa, J. Space, place, and atmosphere: Peripheral perception in existential experience. In Architectural Atmospheres: On the Experience and Politics of Architecture; Borch, C., Ed.; Birkhäuser: Basel, Switzerland, 2014; pp. 18–41. [Google Scholar] [CrossRef]
  2. Semsioglu, S.; Gokce, Y.; Yantac, A.E. Emotionscape: Mediating spatial experience for emotion awareness and sharing. In Proceedings of the Extended Abstracts of the 2018 CHI Conference on Human Factors in Computing Systems, Montreal, QC, Canada, 21–26 April 2018; pp. 1–6. [Google Scholar] [CrossRef]
  3. Lin, S.; Feng, W.; Xiang, J. Color expression of architectural emotion language from the perspective of public art. J. Landsc. Res. 2016, 8, 65. [Google Scholar] [CrossRef]
  4. Bower, I.; Tucker, R.; Enticott, P.G. Impact of built environment design on emotion measured via neurophysiological correlates and subjective indicators: A systematic review. J. Environ. Psychol. 2019, 66, 101344. [Google Scholar] [CrossRef]
  5. Ji, B.; Kang, J.; Kim, C.; Kim, S.; Song, Y.; Yeon, J. The effect of personal characteristics on spatial perception in BIM-based virtual environments: Age, gender, education, and gaming experience. Buildings 2023, 13, 2103. [Google Scholar] [CrossRef]
  6. Bosco, F.J.; Joassart-Marcelli, P. Participatory planning and children’s emotional labor in the production of urban nature. Emot. Space Soc. 2015, 16, 30–40. [Google Scholar] [CrossRef]
  7. Gifford, R. Environmental psychology matters. Annu. Rev. Psychol. 2014, 65, 541–579. [Google Scholar] [CrossRef] [PubMed]
  8. Meng, X.; Zhang, M.; Wang, M. Effects of school indoor visual environment on children’s health outcomes: A systematic review. Health Place 2023, 83, 103021. [Google Scholar] [CrossRef]
  9. Dzhambov, A.M.; Lercher, P.; Vincens, N.; Waye, K.P.; Klatte, M.; Leist, L.; Lachmann, T.; Schreckenberg, D.; Belke, C.; Ristovska, G.; et al. Protective effect of restorative possibilities on cognitive function and mental health in children and adolescents: A scoping review including the role of physical activity. Environ. Res. 2023, 233, 116452. [Google Scholar] [CrossRef]
  10. Wells, N.M. At home with nature: Effects of “greenness” on children’s cognitive functioning. Environ. Behav. 2000, 32, 775–795. [Google Scholar] [CrossRef]
  11. Lau, S.S.Y.; Gou, Z.; Liu, Y. Healthy campus by open space design: Approaches and guidelines. Front. Archit. Res. 2014, 3, 452–467. [Google Scholar] [CrossRef]
  12. Amicone, G.; Petruccelli, I.; De Dominicis, S.; Gherardini, A.; Costantino, V.; Perucchini, P.; Bonaiuto, M. Green breaks: The restorative effect of the school environment’s green areas on children’s cognitive performance. Front. Psychol. 2018, 9, 1579. [Google Scholar] [CrossRef]
  13. Kaplan, S. The restorative benefits of nature: Toward an integrative framework. J. Environ. Psychol. 1995, 15, 169–182. [Google Scholar] [CrossRef]
  14. Finnigan, K.A. Sensory responsive environments: A qualitative study on perceived relationships between outdoor built environments and sensory sensitivities. Land 2024, 13, 636. [Google Scholar] [CrossRef]
  15. Deng, H.; Sulaiman, R.; Ismail, M.A. Enhancing children’s health and well-being through biophilic design in Chinese kindergartens: A systematic literature review. Soc. Sci. Humanit. Open 2024, 10, 100939. [Google Scholar] [CrossRef]
  16. Guo, X.; Tu, X.; Huang, G.; Fang, X.; Kong, L.; Wu, J. Urban greenspace helps ameliorate people’s negative sentiments during the COVID-19 pandemic: The case of Beijing. Build. Environ. 2022, 223, 109449. [Google Scholar] [CrossRef]
  17. Xu, W.; Xu, S.; Shi, R.; Chen, Z.; Lin, Y.; Chen, J. Exploring the impact of university green spaces on students’ perceived restoration and emotional states through audio-visual perception. Ecol. Inform. 2024, 82, 102766. [Google Scholar] [CrossRef]
  18. Wang, F.; Munakata, J. Assessing effects of facade characteristics and visual elements on perceived oppressiveness in high-rise window views via virtual reality. Build. Environ. 2024, 266, 112043. [Google Scholar] [CrossRef]
  19. Gomes, H.; Barrett, S.; Duff, M.; Barnhardt, J.; Ritter, W. The effects of interstimulus interval on event-related indices of attention: An auditory selective attention test of perceptual load theory. Clin. Neurophysiol. 2008, 119, 542–555. [Google Scholar] [CrossRef]
  20. Liu, W.; Wang, B.; Yang, Y.; Mou, N.; Zheng, Y.; Zhang, L.; Yang, T. Cluster analysis of microscopic spatio-temporal patterns of tourists’ movement behaviors in mountainous scenic areas using open GPS-trajectory data. Tour. Manag. 2022, 93, 104614. [Google Scholar] [CrossRef]
  21. Wang, M.; Han, P.; Li, X.; Bao, X.; Huang, J. Continuation and evolution of collective memory manifested in rural public space: Revealed by semi-structured interviews and emotional maps in three migrant villages in Chaihu town. Habitat Int. 2024, 154, 103213. [Google Scholar] [CrossRef]
  22. Le, Q.H.; Kwon, N.; Nguyen, T.H.; Kim, B.; Ahn, Y. Sensing perceived urban stress using space syntactical and urban building density data: A machine learning-based approach. Build. Environ. 2024, 266, 112054. [Google Scholar] [CrossRef]
  23. Göçer, Ö.; Göçer, K.; Başol, A.M.; Kıraç, M.F.; Özbil, A.; Bakovic, M.; Siddiqui, F.P.; Özcan, B. Introduction of a spatio-temporal mapping based POE method for outdoor spaces: Suburban university campus as a case study. Build. Environ. 2018, 145, 125–139. [Google Scholar] [CrossRef]
  24. No, W.; Choi, J.; Kim, Y. How do children move and behave on streets? Vision-based movement behavior analysis using children’s trajectories in urban surveillance systems. Appl. Geogr. 2024, 162, 103170. [Google Scholar] [CrossRef]
  25. Damatac, C.G.; ter Avest, M.J.; Wilderjans, T.F.; De Gucht, V.; Woestenburg, D.H.A.; Landeweerd, L.; Galesloot, T.E.; Geerligs, L.; Homberg, J.R.; Greven, C.U. Exploring sensory processing sensitivity: Relationships with mental and somatic health, interactions with positive and negative environments, and evidence for differential susceptibility. Curr. Res. Behav. Sci. 2025, 8, 100165. [Google Scholar] [CrossRef]
  26. Manahasa, O.; Özsoy, A.; Manahasa, E. Evaluative, inclusive, participatory: Developing a new language with children for school building design. Build. Environ. 2021, 188, 107374. [Google Scholar] [CrossRef]
  27. Gao, W.; Kang, J.; Ma, H.; Wang, C. The effects of environmental sensitivity and noise sensitivity on soundscape evaluation. Build. Environ. 2023, 245, 110945. [Google Scholar] [CrossRef]
  28. Ramkissoon, H.; Van Der Veen, R.; Salaripour, A.; Reihani, Z.S.; Aflaki, A. The impact of sensory experiences on place attachment, place loyalty and civic participation: Evidence from Rasht, Iran. City Cult. Soc. 2024, 38, 100592. [Google Scholar] [CrossRef]
  29. Kaltenborn, B.P.; Bjerke, T. Associations between environmental value orientations and landscape preferences. Landsc. Urban Plan. 2002, 59, 1–11. [Google Scholar] [CrossRef]
  30. Yang, L.; Liu, J.; Albert, C.; Guo, X. Exploring the effects of soundscape perception on place attachment: A comparative study of residents and tourists. Appl. Acoust. 2024, 222, 110048. [Google Scholar] [CrossRef]
  31. Liu, Q.; Wu, Y.; Xiao, Y.; Fu, W.; Zhuo, Z.; Konijnendijk van den Bosch, C.C.; Huang, Q.; Lan, S. More meaningful, more restorative? Linking local landscape characteristics and place attachment to restorative perceptions of urban park visitors. Landsc. Urban Plan. 2020, 197, 103763. [Google Scholar] [CrossRef]
  32. Pretty, G.H.; Chipuer, H.M.; Bramston, P. Sense of place amongst adolescents and adults in two rural Australian towns: The discriminating features of place attachment, sense of community and place dependence in relation to place identity. J. Environ. Psychol. 2003, 23, 273–287. [Google Scholar] [CrossRef]
  33. Scheller, D.A.; Sterr, K.; Humpe, A.; Mess, F.; Bachner, J. Physical activity through place attachment: Understanding perceptions of children and adolescents on urban places by using photovoice and walking interviews. Health Place 2024, 90, 103361. [Google Scholar] [CrossRef]
  34. Fung, W.K.; Chung, K.K.H. Longitudinal association between children’s mastery motivation and cognitive school readiness: Executive functioning and social-emotional competence as potential mediators. J. Exp. Child Psychol. 2023, 234, 105712. [Google Scholar] [CrossRef] [PubMed]
  35. Yang, F.; Xia, Z. Perceived discrimination and academic self-concept among left-behind children in China: The role of school belonging and classroom composition. Child. Youth Serv. Rev. 2023, 155, 107294. [Google Scholar] [CrossRef]
  36. Khotbehsara, E.M.; Somasundaraswaran, K.; Kolbe-Alexander, T.; Yu, R. The influence of spatial configuration on pedestrian movement behaviour in commercial streets of low-density cities. Ain Shams Eng. J. 2025, 16, 103184. [Google Scholar] [CrossRef]
  37. Zhang, J.; Liu, S.; Liu, K.; Bian, F. How does campus-scape influence university students’ restorative experiences: Evidences from simultaneously collected physiological and psychological data. Urban For. Urban Green. 2025, 107, 128779. [Google Scholar] [CrossRef]
  38. Liu, Q.; Luo, S.; Shen, Y.; Zhu, Z.; Yao, X.; Li, Q.; Tarin, M.W.K.; Zheng, J.; Zhuo, Z. Relationships between students’ demographic characteristics, perceived naturalness and patterns of use associated with campus green space, and self-rated restoration and health. Urban For. Urban Green. 2022, 68, 127474. [Google Scholar] [CrossRef]
  39. Li, H.; Du, J.; Chow, D. Perceived environmental factors and students’ mental wellbeing in outdoor public spaces of university campuses: A systematic scoping review. Build. Environ. 2024, 265, 112023. [Google Scholar] [CrossRef]
  40. Bianchi, E.; Bencharit, L.Z.; Murnane, E.L.; Altaf, B.; Douglas, I.P.; Landay, J.A.; Billington, S.L. Effects of architectural interventions on psychological, cognitive, social, and pro-environmental aspects of occupant well-being: Results from an immersive online study. Build. Environ. 2024, 253, 111293. [Google Scholar] [CrossRef]
  41. Sun, K.; Li, Z.; Zheng, S.; Qu, H. Quantifying environmental characteristics on psychophysiological restorative benefits of campus window views. Build. Environ. 2024, 262, 111822. [Google Scholar] [CrossRef]
  42. Lin, M.; Lu, W.; Li, N.; Geng, W. Exploring the relationship between library reading space height and user perception: A physiological and subjective analysis. Build. Environ. 2024, 253, 111307. [Google Scholar] [CrossRef]
  43. Barrett, P.; Davies, F.; Zhang, Y.; Barrett, L. The impact of classroom design on pupils’ learning: Final results of a holistic, multi-level analysis. Build. Environ. 2015, 89, 118–133. [Google Scholar] [CrossRef]
  44. Vivion, M.; Ftaïta, M.; Guida, A.; Mathy, F. Processing order in short-term memory is spatially biased in children. J. Exp. Child Psychol. 2025, 252, 106171. [Google Scholar] [CrossRef] [PubMed]
  45. Barrett, P.; Zhang, Y.; Moffat, J.; Kobbacy, K. A holistic, multi-level analysis identifying the impact of classroom design on pupils’ learning. Build. Environ. 2013, 59, 678–689. [Google Scholar] [CrossRef]
  46. Harris, D.J. A quantitative approach to the assessment of the environmental impact of building materials. Build. Environ. 1999, 34, 751–758. [Google Scholar] [CrossRef]
  47. Severcan, Y.C.; Torun, A.O.; Defeyter, M.A.; Bingol, H.; Akin, I.Z. Associations of children’s mental wellbeing and the urban form characteristics of their everyday places. Cities 2025, 160, 105832. [Google Scholar] [CrossRef]
  48. Van Den Bogerd, N.; Dijkstra, S.C.; Seidell, J.C.; Maas, J. Greenery in the university environment: Students’ preferences and perceived restoration likelihood. PLoS ONE 2018, 13, e0192429. [Google Scholar] [CrossRef]
  49. Liu, Y.; Zhang, J.; Liu, C.; Yang, Y. A review of attention restoration theory: Implications for designing restorative environments. Sustainability 2024, 16, 3639. [Google Scholar] [CrossRef]
  50. McLenon, J.; Rogers, M.A. The fear of needles: A systematic review and meta-analysis. J. Adv. Nurs. 2019, 75, 30–42. [Google Scholar] [CrossRef]
  51. Allah Yar, M.; Kazemi, F. The role of dish gardens on the physical and neuropsychological improvement of hospitalized children. Urban For. Urban Green. 2020, 53, 126713. [Google Scholar] [CrossRef]
  52. Zhao, G.; Xiao, L.-R.; Chen, Y.-H.; Zhang, M.; Peng, K.-W.; Wu, H.-M. Association between physical activity and mental health problems among children and adolescents: A moderated mediation model of emotion regulation and gender. J. Affect. Disord. 2025, 369, 489–498. [Google Scholar] [CrossRef]
  53. Jansson, M. Children’s perspectives on public playgrounds in two Swedish communities. Child. Youth Environ. 2008, 18, 88–109. Available online: https://pub.epsilon.slu.se/9410/7/Jansson_M_130201.pdf (accessed on 3 March 2023).
  54. Cui, J.; Meng, X.; Qi, S.; Fan, J.; Yu, W.; Liu, H.; Wang, X.; Zhang, Y. The impact of school activity space layout on children’s physical activity levels during recess: An agent-based model computational approach. Build. Environ. 2025, 271, 112585. [Google Scholar] [CrossRef]
  55. Ding, P.; Carstensen, T.A.; Jørgensen, G. Exploring the inclusion of children from a spatial perspective: An analytical framework of the correlation between physical environment and children’s inclusion in urban public spaces. Cities 2024, 153, 105293. [Google Scholar] [CrossRef]
  56. Ulrich, R.S. View through a window may influence recovery from surgery. Science 1984, 224, 420–421. [Google Scholar] [CrossRef]
  57. Lopez, R.P.; Hynes, H.P. Obesity, physical activity, and the urban environment: Public health research needs. Environ. Health 2006, 5, 25. [Google Scholar] [CrossRef]
  58. Li, S.; Huang, C. Using convolutional neural networks for image semantic segmentation and object detection. Syst. Soft Comput. 2024, 6, 200172. [Google Scholar] [CrossRef]
  59. Song, C.; Wu, H.; Ma, X. Semantic-guided modeling of spatial relation and object co-occurrence for indoor scene recognition. Expert Syst. Appl. 2025, 270, 126415. [Google Scholar] [CrossRef]
  60. Huang, B.; Wang, L.; Chen, T.; Kong, G.; Shuai, R.; Chen, P. The influence of interface proportions on visual guidance perception in node spaces. Build. Environ. 2025, 272, 112661. [Google Scholar] [CrossRef]
  61. van Heezik, Y.; Freeman, C.; Falloon, A.; Buttery, Y.; Heyzer, A. Relationships between childhood experience of nature and green/blue space use, landscape preferences, connection with nature and pro-environmental behavior. Landsc. Urban Plan. 2021, 213, 104135. [Google Scholar] [CrossRef]
  62. Diachenko, I.; Kalishchuk, S.; Zhylin, M.; Kyyko, A.; Volkova, Y. Color education: A study on methods of influence on memory. Heliyon 2022, 8, e11607. [Google Scholar] [CrossRef]
  63. Wang, Y.; Yuan, L. Analysis of the campus design based on the color idea of landscape space. In Civil Engineering and Urban Planning III; CRC Press: Boca Raton, FL, USA, 2014; pp. 59–64. [Google Scholar] [CrossRef]
  64. Yildirim, K.; Akalin-Baskaya, A.; Hidayetoglu, M.L. Effects of indoor color on mood and cognitive performance. Build. Environ. 2007, 42, 3233–3240. [Google Scholar] [CrossRef]
  65. Shimatani, K.; Nakayama, Y.; Takaguchi, K.; Iwayama, R.; Yoda-Tsumura, K.; Nakaoka, H.; Mori, C.; Suzuki, N. Relationship between living rooms with void spaces or partially high ceilings and psychological well-being: A cross-sectional study in Japan. Build. Environ. 2024, 258, 111596. [Google Scholar] [CrossRef]
  66. Mehta, R.; Zhu, R. Creating when you have less: The impact of resource scarcity on product use creativity. J. Consum. Res. 2016, 42, 767–782. [Google Scholar] [CrossRef]
  67. Gómez Sirvent, J.L.; Fernández-Sotos, D.; Sánchez-Reolid, R.; de la Rosa López, F.; Fernández-Sotos, A.; Fernández-Caballero, A. Pre-occupancy evaluation of a virtual music school classroom: Influence of color and type of lighting on music performers. Build. Environ. 2023, 246, 110989. [Google Scholar] [CrossRef]
  68. Li, J.; Chen, S.; Xu, H.; Kang, J. Effects of implanted wood components on environmental restorative quality of indoor informal learning spaces in college. Build. Environ. 2023, 245, 110890. [Google Scholar] [CrossRef]
  69. Fan, T.; Tang, X.; Li, K. Image clustering algorithm and psychological perception in historical building colour rating research: A case study of Guangzhou, China. Front. Archit. Res. 2025, 14, 1415–1435. [Google Scholar] [CrossRef]
  70. Zhang, R.; Huang, Q.; Peng, Z.; Zhang, X.; Shang, L.; Yang, C. Evaluating the impact of elementary school urban neighborhood color on children’s mentalization of emotions through multi-source data. Buildings 2024, 14, 3128. [Google Scholar] [CrossRef]
  71. Duan, Y.; Wu, J. Sport tourist perceptions of destination image and revisit intentions: An adaption of Mehrabian-Russell’s environmental psychology model. Heliyon 2024, 10, e31810. [Google Scholar] [CrossRef]
  72. Pan, S. Emotional analysis of broadcasting and hosting speech by integrating grid PSO-SVR and PAD models. Int. J. Cogn. Comput. Eng. 2025, 6, 55–64. [Google Scholar] [CrossRef]
  73. Hideg, É.; Mihók, B.; Gáspár, J.; Schmidt, P.; Márton, A.; Báldi, A. Assessment in horizon scanning by various stakeholder groups using Osgood’s semantic differential scale—A methodological development. Futures 2021, 126, 102677. [Google Scholar] [CrossRef]
  74. Tzeng, O.C.S.; Landis, D.; Tzeng, D.Y.; Charles, E. Osgood’s continuing contributions to intercultural communication and far beyond! Int. J. Intercult. Relat. 2012, 36, 832–842. [Google Scholar] [CrossRef]
  75. Osgood, C.E. Dimensionality of the semantic space for communication via facial expressions. Scand. J. Psychol. 1966, 7, 1–30. [Google Scholar] [CrossRef]
  76. Taverner, J.; Vivancos, E.; Botti, V. A fuzzy appraisal model for affective agents adapted to cultural environments using the pleasure and arousal dimensions. Inf. Sci. 2021, 546, 74–86. [Google Scholar] [CrossRef]
  77. Mehrabian, A. Framework for a comprehensive description and measurement of emotional states. Genet. Soc. Gen. Psychol. Monogr. 1995, 121, 339–361. [Google Scholar] [CrossRef] [PubMed]
  78. Mehrabian, A. Pleasure-arousal-dominance: A general framework for describing and measuring individual differences in temperament. Curr. Psychol. 1996, 14, 261–292. [Google Scholar] [CrossRef]
  79. Floyd, M.F. Pleasure, arousal, and dominance: Exploring affective determinants of recreation satisfaction. Leis. Sci. 1997, 19, 83–96. [Google Scholar] [CrossRef]
  80. Preißler, L.; Keck, J.; Krüger, B.; Munzert, J.; Schwarzer, G. Recognition of emotional body language from dyadic and monadic point-light displays in 5-year-old children and adults. J. Exp. Child Psychol. 2023, 235, 105713. [Google Scholar] [CrossRef]
  81. Day, T.N.; Mazefsky, C.A.; Yu, L.; Zeglen, K.N.; Neece, C.L.; Pilkonis, P.A. The Emotion Dysregulation Inventory−Young Child: Psychometric properties and item response theory calibration in 2- to 5-year-olds. J. Am. Acad. Child Adolesc. Psychiatry 2024, 63, 52–64. [Google Scholar] [CrossRef]
  82. Neves, L.; Martins, M.; Correia, A.I.; Castro, S.L.; Schellenberg, E.G.; Lima, C.F. Does music training improve emotion recognition and cognitive abilities? Longitudinal and correlational evidence from children. Cognition 2025, 259, 106102. [Google Scholar] [CrossRef]
  83. England-Mason, G.; Andrews, K.; Atkinson, L.; Gonzalez, A. Emotion socialization parenting interventions targeting emotional competence in young children: A systematic review and meta-analysis of randomized controlled trials. Clin. Psychol. Rev. 2023, 100, 102252. [Google Scholar] [CrossRef]
  84. Bacso, S.A.; Nilsen, E.S. Children’s use of verbal and nonverbal feedback during communicative repair: Associations with executive functioning and emotion knowledge. Cogn. Dev. 2022, 63, 101199. [Google Scholar] [CrossRef]
  85. Galler, M.; Grendstad, Å.R.; Ares, G.; Varela, P. Capturing food-elicited emotions: Facial decoding of children’s implicit and explicit responses to tasted samples. Food Qual. Prefer. 2022, 99, 104551. [Google Scholar] [CrossRef]
  86. Lercher, P.; Dzhambov, A.M.; Waye, K.P. Environmental perceptions, self-regulation, and coping with noise mediate the associations between children’s physical environment and sleep and mental health problems. Environ. Res. 2025, 264 Pt 2, 120414. [Google Scholar] [CrossRef] [PubMed]
  87. Gitelman, V.; Levi, S.; Carmel, R.; Korchatov, A.; Hakkert, S. Exploring patterns of child pedestrian behaviors at urban intersections. Accid. Anal. Prev. 2019, 122, 36–47. [Google Scholar] [CrossRef]
  88. Ibes, D.C.; Forestell, C.A. The role of campus greenspace and meditation on college students’ mood disturbance. J. Am. Coll. Health 2022, 70, 99–106. [Google Scholar] [CrossRef]
  89. Günaydın, A.S.; Yücekaya, M. An investigation of sustainable transportation model in campus areas with space syntax method. ICONARP Int. J. Archit. Plan. 2020, 8, 262–281. [Google Scholar] [CrossRef]
  90. Meng, X.; Zhang, M. Effects of classroom design characteristics on children’s physiological and psychological responses: A virtual reality experiment. Build. Environ. 2025, 267 Pt B, 112274. [Google Scholar] [CrossRef]
  91. Gonzalez, D.; Rueda-Plata, D.; Acevedo, A.B.; Duque, J.C.; Ramos-Pollán, R.; Betancourt, A.; García, S. Automatic detection of building typology using deep learning methods on street level images. Build. Environ. 2020, 177, 106805. [Google Scholar] [CrossRef]
  92. Shen, X.; Wu, Y.; Liu, F.; Kang, J. Spatial adaptation in children with autism—A study based on behavioural dynamic video data. J. Build. Eng. 2024, 98, 111228. [Google Scholar] [CrossRef]
  93. Lan, B.; Yu, Z.J.; Zhou, P.; Huang, G. Optimal zoning for building zonal model of large-scale indoor space. Build. Environ. 2022, 225, 109669. [Google Scholar] [CrossRef]
  94. Meng, Y.; Wang, J.; Xi, C.; Han, L.; Feng, Z.; Cao, S.-J. Investigation of heat stress on urban roadways for commuting children and mitigation strategies from the perspective of urban design. Urban Clim. 2023, 49, 101564. [Google Scholar] [CrossRef]
  95. Griffith, Z.M.; Polet, J.; Lintunen, T.; Hamilton, K.; Hagger, M.S. Social cognition, personality and social-political correlates of health behaviors: Application of an integrated theoretical model. Soc. Sci. Med. 2024, 347, 116779. [Google Scholar] [CrossRef] [PubMed]
  96. Black, J.T.; Shakir, M.Z. Emotion on the edge: An evaluation of feature representations and machine learning models. Nat. Lang. Process. J. 2025, 10, 100127. [Google Scholar] [CrossRef]
  97. Calderon-Uribe, S.; Morales-Hernandez, L.A.; Guzman-Sandoval, V.M.; Dominguez-Trejo, B.; Cruz-Albarran, I.A. Emotion detection based on infrared thermography: A review of machine learning and deep learning algorithms. Infrared Phys. Technol. 2025, 145, 105669. [Google Scholar] [CrossRef]
  98. Prakash, A.; Poulose, A. Electroencephalogram-based emotion recognition: A comparative analysis of supervised machine learning algorithms. Data Sci. Manag. 2025, 8, 342–360. [Google Scholar] [CrossRef]
  99. Esfahani, S.H.N.; Adda, M. Classical machine learning and large models for text-based emotion recognition. Procedia Comput. Sci. 2024, 241, 77–84. [Google Scholar] [CrossRef]
  100. Zhao, Y.; Zhang, Y. Research on implicit emotion recognition and classification in literary works in the context of machine learning. Alex. Eng. J. 2025, 115, 577–584. [Google Scholar] [CrossRef]
  101. Subasi, A.; Qaisar, S.M. EEG-based emotion recognition using AR burg and ensemble machine learning models. In Artificial Intelligence Applications in Healthcare and Medicine; Subasi, A., Qaisar, S.M., Nisar, H., Eds.; Academic Press: Cambridge, MA, USA, 2025; pp. 303–329. [Google Scholar] [CrossRef]
  102. Bisogni, C.; Cimmino, L.; De Marsico, M.; Hao, F.; Narducci, F. Emotion recognition at a distance: The robustness of machine learning based on hand-crafted facial features vs deep learning models. Image Vis. Comput. 2023, 136, 104724. [Google Scholar] [CrossRef]
  103. Garg, M.; Saxena, C. Emotion detection from text data using machine learning for human behavior analysis. In Computational Intelligence Methods for Sentiment Analysis in Natural Language Processing Applications; Hemanth, D.J., Ed.; Morgan Kaufmann: Burlington, MA, USA, 2024; pp. 129–144. [Google Scholar] [CrossRef]
Figure 1. Research framework.
Figure 1. Research framework.
Buildings 15 04281 g001
Figure 2. SAM (Segment Anything Model) model structure.
Figure 2. SAM (Segment Anything Model) model structure.
Buildings 15 04281 g002
Figure 3. Importance of emotional characteristics of spatial elements.
Figure 3. Importance of emotional characteristics of spatial elements.
Buildings 15 04281 g003
Figure 4. Utility of children’s emotional indicators by gender.
Figure 4. Utility of children’s emotional indicators by gender.
Buildings 15 04281 g004
Figure 5. Utility of children’s emotional indicators by age.
Figure 5. Utility of children’s emotional indicators by age.
Buildings 15 04281 g005
Table 1. Indicators of Spatial Color in Elementary School Campuses.
Table 1. Indicators of Spatial Color in Elementary School Campuses.
Target LayerGuideline LayerFactor Layer
Elementary school campus spaceColor richness B1Number of colors C1
Primary color ratio C2
Color Harmony Type C3
Visual Impact B2Hue Contrast C4
Saturation Contrast C5
Brightness Contrast C6
Color Performance B3Hue Index C7
Saturation Index C8
Brightness Index C9
Color warmth and coolness C10
Table 2. Composition of types of children’s campus indoor and outdoor space behaviors.
Table 2. Composition of types of children’s campus indoor and outdoor space behaviors.
Static BehaviorDynamic BehaviorIndividual BehaviorGroup BehaviorTotal
TotalPercentageTotalPercentageTotalPercentageTotalPercentage
Outside space17341%23951%20349%20944%412
Internal spaceLearning Zone12329%368%7017%8919%159
Recreational areas338%6414%318%6614%97
Active Area9021%12627%10826%10823%216
Total419465412472884
Table 3. Distribution of children’s emotions in different elementary school campus spaces.
Table 3. Distribution of children’s emotions in different elementary school campus spaces.
Distribution of Space
ArousalBuildings 15 04281 i001
DominanceBuildings 15 04281 i002
PleasureBuildings 15 04281 i003
Table 4. Correlation between color indicators and spatial elements.
Table 4. Correlation between color indicators and spatial elements.
Number of
Colors C1
Main Color
Proportion C2
Color
Harmony Type C3
Hue
Contrast C4
Saturation Contrast C5Luminance Contrast C6Hue Index C7Saturation Index C8Brightness Index C9Color Warmth and Coolness C10
Table and chair appliance0.164 (0.000 ***)−0.035 (0.002 ***)−0.049 (0.000 ***)−0.233 (0.000 ***)−0.021 (0.071 *)−0.08 (0.000 ***)−0.238 (0.000 ***)−0.056 (0.000 ***)0.219 (0.000 ***)−0.077 (0.000 ***)
Cabinet furniture−0.128 (0.000 ***)−0.075 (0.000 ***)0.039 (0.001 ***)−0.115 (0.000 ***)−0.046 (0.000 ***)−0.178 (0.000 ***)−0.186 (0.000 ***)0.207 (0.000 ***)0.258 (0.000 ***)0.244 (0.000 ***)
Floor paving−0.264 (0.000 ***)−0.125 (0.000 ***)−0.414 (0.000 ***)−0.286 (0.000 ***)−0.366 (0.000 ***)−0.227 (0.000 ***)−0.407 (0.000 ***)−0.385 (0.000 ***)0.04 (0.000 ***)0.106 (0.000 ***)
ceiling−0.02 (0.080 *)0.344 (0.000 ***)−0.177 (0.000 ***)−0.248 (0.000 ***)−0.192 (0.000 ***)−0.072 (0.000 ***)−0.293 (0.000 ***)−0.234 (0.000 ***)−0.019 (0.100 *)0.119 (0.000 ***)
Enclose the wall elevation−0.122 (0.000 ***)−0.119 (0.000 ***)−0.419 (0.000 ***)−0.377 (0.000 ***)−0.14 (0.000 ***)−0.258 (0.000 ***)−0.449 (0.000 ***)−0.051 (0.000 ***)0.261 (0.000 ***)0.018 (0.114)
Door and window construction−0.201 (0.000 ***)−0.02 (0.076 *)−0.338 (0.000 ***)−0.239 (0.000 ***)−0.271 (0.000 ***)−0.203 (0.000 ***)−0.297 (0.000 ***)−0.228 (0.000 ***)0.179 (0.000 ***)0.025 (0.027 **)
Facilities and decor−0.017 (0.141)0.318 (0.000 ***)−0.052 (0.000 ***)−0.209 (0.000 ***)0.149 (0.000 ***)0.122 (0.000 ***)−0.225 (0.000 ***)0.362 (0.000 ***)0.011 (0.334)0.306 (0.000 ***)
Indoor greening0.524 (0.000 ***)0.167 (0.000 ***)0.324 (0.000 ***)−0.079 (0.000 ***)0.463 (0.000 ***)−0.025 (0.025 **)0.061 (0.000 ***)−0.002 (0.891)0.091 (0.000 ***)0.064 (0.000 ***)
Building elevation−0.09 (0.000 ***)0.091 (0.000 ***)0.009 (0.426)−0.016 (0.157)0.005 (0.661)−0.12 (0.000 ***)−0.003 (0.784)−0.054 (0.000 ***)0.038 (0.001 ***)0.204 (0.000 ***)
Floor paving−0.12 (0.000 ***)−0.145 (0.000 ***)0.253 (0.000 ***)0.166 (0.000 ***)0.208 (0.000 ***)0.08 (0.000 ***)0.168 (0.000 ***)−0.023 (0.041 **)−0.046 (0.000 ***)0.047 (0.000 ***)
Road path0.133 (0.000 ***)0.112 (0.000 ***)0.255 (0.000 ***)0.11 (0.000 ***)−0.066 (0.000 ***)0.025 (0.030 **)0.49 (0.000 ***)−0.081 (0.000 ***)−0.24 (0.000 ***)−0.17 (0.000 ***)
playground0.049 (0.000 ***)−0.129 (0.000 ***)0.218 (0.000 ***)0.237 (0.000 ***)0.32 (0.000 ***)0.026 (0.023 **)0.169 (0.000 ***)0.011 (0.336)−0.094 (0.000 ***)0.025 (0.026 **)
Building elevation0.34 (0.000 ***)0.094 (0.000 ***)−0.025 (0.031 **)0.29 (0.000 ***)0.063 (0.000 ***)0.407 (0.000 ***)0.311 (0.000 ***)0.085 (0.000 ***)−0.115 (0.000 ***)−0.27 (0.000 ***)
Enclosure construction−0.084 (0.000 ***)−0.165 (0.000 ***)0.362 (0.000 ***)0.357 (0.000 ***)0.063 (0.000 ***)−0.033 (0.003 ***)0.406 (0.000 ***)−0.005 (0.664)−0.188 (0.000 ***)−0.037 (0.001 ***)
Facilities guide system−0.019 (0.102)−0.083 (0.000 ***)−0.132 (0.000 ***)0.075 (0.000 ***)−0.028 (0.013 **)0.511 (0.000 ***)0.051 (0.000 ***)0.058 (0.000 ***)0.046 (0.000 ***)−0.148 (0.000 ***)
Green landscape−0.05 (0.000 ***)−0.08 (0.000 ***)0.301 (0.000 ***)0.334 (0.000 ***)−0.062 (0.000 ***)0.296 (0.000 ***)0.523 (0.000 ***)0.058 (0.000 ***)−0.267 (0.000 ***)−0.289 (0.000 ***)
Sky view0.158 (0.000 ***)−0.074 (0.000 ***)0.523 (0.000 ***)0.566 (0.000 ***)0.329 (0.000 ***)0.196 (0.000 ***)0.49 (0.000 ***)0.006 (0.600)−0.087 (0.000 ***)−0.225 (0.000 ***)
Note: ***, **, * represent 1%, 5%, and 10% significance levels, respectively.
Table 5. Behavior Map of Children’s Campus Exterior Building Space.
Table 5. Behavior Map of Children’s Campus Exterior Building Space.
SpaceBehavioral MappingActivity Frequency/Spatial Elements
Building facadeInner court areaBuildings 15 04281 i013Buildings 15 04281 i014
Buildings 15 04281 i015Buildings 15 04281 i016
Motor areaBuildings 15 04281 i017Buildings 15 04281 i018
Buildings 15 04281 i019Buildings 15 04281 i020
Activity areaBuildings 15 04281 i021Buildings 15 04281 i022
Buildings 15 04281 i023Buildings 15 04281 i024
Overhead areaBuildings 15 04281 i025Buildings 15 04281 i026
Buildings 15 04281 i027Buildings 15 04281 i028
Outdoor characteristics
district
Buildings 15 04281 i029Buildings 15 04281 i030
Buildings 15 04281 i031Buildings 15 04281 i032
Enclosed memberBuildings 15 04281 i033Buildings 15 04281 i034
Buildings 15 04281 i035Buildings 15 04281 i036
Motion pathBuildings 15 04281 i037Buildings 15 04281 i038
Buildings 15 04281 i039Buildings 15 04281 i040
Public sports facilitiesBuildings 15 04281 i041Buildings 15 04281 i042
Buildings 15 04281 i043Buildings 15 04281 i044
Unsteady embellishmentBuildings 15 04281 i045Buildings 15 04281 i046
Buildings 15 04281 i047Buildings 15 04281 i048
Table 6. Spatial Behavior Map of Learning Area in Children’s Campus Building.
Table 6. Spatial Behavior Map of Learning Area in Children’s Campus Building.
SpaceBehavioral MapsActivity Frequency/Spatial Features
Learning areaLibraryBuildings 15 04281 i049Buildings 15 04281 i050
Buildings 15 04281 i051Buildings 15 04281 i052
Study classroomBuildings 15 04281 i053Buildings 15 04281 i054
Buildings 15 04281 i055Buildings 15 04281 i056
Multifunctional classroomBuildings 15 04281 i057Buildings 15 04281 i058
Buildings 15 04281 i059Buildings 15 04281 i060
Lecture hallBuildings 15 04281 i061Buildings 15 04281 i062
Buildings 15 04281 i063Buildings 15 04281 i064
Table 7. Spatial Behavior Map of Living and Leisure Areas inside Children’s Campus Buildings.
Table 7. Spatial Behavior Map of Living and Leisure Areas inside Children’s Campus Buildings.
Behavioral MapBehavior MapActivity FrequencySpatial Elements
Leisure areaRestroomsBuildings 15 04281 i065Buildings 15 04281 i066Buildings 15 04281 i067
Buildings 15 04281 i068
CafeteriaBuildings 15 04281 i069Buildings 15 04281 i070Buildings 15 04281 i071
Buildings 15 04281 i072
Atrium spaceBuildings 15 04281 i073Buildings 15 04281 i074Buildings 15 04281 i075
Buildings 15 04281 i076
Table 8. Spatial Behavior Map of Activity Areas in Children’s Campus Buildings.
Table 8. Spatial Behavior Map of Activity Areas in Children’s Campus Buildings.
Behavioral MapBehavior MapActivity Frequency/Spatial Elements
Activity areaPlantationBuildings 15 04281 i077Buildings 15 04281 i078
Buildings 15 04281 i079Buildings 15 04281 i080
PlaygroundBuildings 15 04281 i081Buildings 15 04281 i082
Buildings 15 04281 i083Buildings 15 04281 i084
Corridor spaceBuildings 15 04281 i085Buildings 15 04281 i086
Buildings 15 04281 i087Buildings 15 04281 i088
StairwellsBuildings 15 04281 i089Buildings 15 04281 i090
Buildings 15 04281 i091Buildings 15 04281 i092
Shared Activity SpaceBuildings 15 04281 i093Buildings 15 04281 i094
Buildings 15 04281 i095Buildings 15 04281 i096
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

Zhang, R.; Li, B.; Huang, Q.; Peng, Z.; Xu, Y.; Tang, L.; Ouyang, Z.; Zhang, X.; Shang, L. Spatial Optimization of Primary School Campuses from the Perspective of Children’s Emotional Behavior: A Deep Learning and Machine Learning Approach. Buildings 2025, 15, 4281. https://doi.org/10.3390/buildings15234281

AMA Style

Zhang R, Li B, Huang Q, Peng Z, Xu Y, Tang L, Ouyang Z, Zhang X, Shang L. Spatial Optimization of Primary School Campuses from the Perspective of Children’s Emotional Behavior: A Deep Learning and Machine Learning Approach. Buildings. 2025; 15(23):4281. https://doi.org/10.3390/buildings15234281

Chicago/Turabian Style

Zhang, Ruiying, Binghuan Li, Qian Huang, Zhimou Peng, Yixun Xu, Li Tang, Zhiyue Ouyang, Xinyue Zhang, and Lan Shang. 2025. "Spatial Optimization of Primary School Campuses from the Perspective of Children’s Emotional Behavior: A Deep Learning and Machine Learning Approach" Buildings 15, no. 23: 4281. https://doi.org/10.3390/buildings15234281

APA Style

Zhang, R., Li, B., Huang, Q., Peng, Z., Xu, Y., Tang, L., Ouyang, Z., Zhang, X., & Shang, L. (2025). Spatial Optimization of Primary School Campuses from the Perspective of Children’s Emotional Behavior: A Deep Learning and Machine Learning Approach. Buildings, 15(23), 4281. https://doi.org/10.3390/buildings15234281

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