Biophilic Value Structure of Forests Perceived by University Students: A Text Mining Approach

Article information

J. People Plants Environ. 2026;29(3):409-422
Publication date (electronic) : 2026 June 30
doi : https://doi.org/10.11628/ksppe.2026.29.3.409
1Postdoctoral Researcher, Forest Human Service Division, Future Forest Strategy Department, National Institute of Forest Science, Seoul, Republic of Korea
2Researcher, Forest Human Service Division, Future Forest Strategy Department, National Institute of Forest Science, Seoul, Republic of Korea
3Doctoral Student, Department of Urban Planning and Real Estate, Graduate School, Kwangwoon University, Seoul, Republic of Korea
*Corresponding author: Seonghak Kim, ksh0615@korea.kr, https://orcid.org/0000-0002-0847-6471
First author: Jinhae Chae, cstarsea@daum.net, https://orcid.org/0000-0002-6334-7613
This study was supported by the National Institute of Forest Science (NIFoS) Research Project (FM0500-2023-01-2025), Korea Forest Service.
Received 2026 January 2; Revised 2026 February 6; Accepted 2026 April 13.

Abstract

Background and objective

With growing interest in intergenerational justice in the context of climate change, increasing attention has been directed toward how forests are perceived as resources that support resilience. This study investigates university students’ perceptions of forest values and explores the functional and structural significance of forests as biophilic infrastructure.

Methods

Unstructured written responses were collected from university students and analyzed using text-mining techniques, including keyword frequency analysis, cluster analysis, centrality analysis, and sentiment analysis.

Results

The results confirmed the presence of biophilic design patterns and their associated effects in university students’ perceptions of forests. Cluster analysis classified these perceptions into six distinct value types: Attraction, Reason, Affection, Spirituality, Utilitarian, and Symbolism. Centrality analysis revealed that Attraction and Reason emerged as core values associated with “Nature in the Space,” reflecting sensory and cognitive engagement with forests. Spirituality and Affection functioned as key complementary values linked to “Nature of the Space,” supporting emotional bonding and psychological restoration. Meanwhile, Utilitarian and Symbolism were identified as peripheral values associated with functional use and cultural interpretation. Sentiment analysis revealed a high proportion of neutral expressions, indicating that forests are embedded in everyday environments. At the same time, the predominance of positive expressions, particularly those related to diversity and richness, suggests that forests facilitate positive experiential responses. These findings demonstrate that forests function as experiential environments in which multiple biophilic values are hierarchically structured. The identified value structure reflects how forests are interpreted and experienced, highlighting their relevance to intergenerational justice in the context of climate change. Understanding these perceptions provides important insights into how nature-based environments can contribute to fostering long-term environmental awareness and resilience.

Conclusion

This study is significant in that it empirically conceptualizes biophilia as a structured value system and frames urban forests as environments in which human experiences and meanings are formed. Although the findings are limited to a university student population, they provide a foundation for future research comparing different generations and types of urban nature.

Introduction

Research Background and Purpose

In contemporary cities, where climate change and urban densification are rapidly intensifying, the urban environment is no longer limited to physical infrastructure serving functional purposes. Rather, it has evolved into an environment that directly shapes citizens’ emotions and everyday experiences. In particular, with the post-pandemic increase in demand for relaxation, recovery, and emotional stability, growing attention has been directed toward the role of natural environments within cities, especially forests. Whereas forests have traditionally been discussed primarily in terms of their physical functions—such as climate regulation, water purification, and biodiversity conservation—their significance has recently expanded to encompass their role as “urban experiential spaces” in which citizens’ emotions, perceptions, and everyday experiences accumulate.

The experiential characteristics emerging through human-forest interactions are grounded in the concept of biophilia, which posits that humans possess an innate tendency to seek connections with nature (Wilson, 1986). Building upon this concept, the notion of biophilic cities was developed to interpret the experiential qualities manifested within urban settings. This perspective reconceptualizes cities as integrated environments composed of natural elements and human life, evolving into an attempt to structurally incorporate nature-based experiences throughout urban spaces (Beatley and Newman, 2013). Within biophilic cities, forests are understood not merely as green elements, but as integral components of urban ecological networks and as key spaces mediating everyday encounters with nature. From an urban design perspective, forests function beyond their physical form as design elements that shape citizens’ emotions, behaviors, and interpretations of meaning. Kellert and Wilson (1993) presented the human-nature relationship—the fundamental principle through which such design elements operate—as eight value dimensions, including attraction, affection, reason, symbolism, and spirituality. These biophilic value dimensions do not exist independently; rather, they constitute an interconnected cognitive framework that manifests differently according to individual experiences and social contexts. Forests may therefore be regarded as representative natural environments in which these values are concretely experienced and interpreted. This phenomenon is consistent with Kaplan and Kaplan’s (1989) Attention Restoration Theory (ART), which posits that natural environments facilitate emotional stability and cognitive restoration. Supporting this theoretical perspective, previous studies have confirmed that urban forests serve as environments shaping psychological and emotional experiences beyond their role as mere physical spaces (Jaszczak et al., 2020). Similarly, research in South Korea has interpreted forests as cultural spaces onto which memories, symbols, and social meanings are projected (Chae, 2016; Chae et al., 2025). Furthermore, from the perspective of biophilic design, prior research has shown that the spatial integration of natural elements is closely associated with enhanced psychological comfort, improved well-being, and strengthened urban resilience (Söderlund and Newman, 2015; Hung and Chang, 2022; Yang and Lee, 2023; Lee, 2023). Collectively, these studies suggest that forests function not merely as environmental backdrops, but as active design elements that influence human emotions, perceptions, and behaviors.

However, to examine these theoretical characteristics, existing research has primarily focused on analyzing physical environmental attributes or the effects of forest use. Consequently, quantitative and structural investigations into the value systems through which individuals perceive forests—as well as the relational networks and structures formed among such values—remain relatively limited. Climate change is increasingly recognized not simply as an environmental issue, but as a matter of intergenerational justice directly linked to the living conditions of future generations (Brown Weiss, 1989; Mintzer and Michel, 2001; Weston, 2008). In this regard, understanding how future generations perceive and interpret the environment and nature is of critical importance (Soga et al., 2016). Forests, in particular, function as nature-based resources that contribute to climate change mitigation and resilience enhancement (Seddon et al., 2020). Accordingly, identifying the values through which future generations perceive forests may provide important insights into future directions of environmental adaptation.

To empirically examine biophilic perceptions of forests, this study focuses on university students. Since forests are sustainable resources that have accumulated over a long period, with their formation and value realization not being temporary, understanding the spatial perceptions of university students among future generations is essential for responding to continuous urban environmental changes driven by climate change and urban densification. Specifically, university students not only serve as core agents shaping social trends and mainstream culture (Min, 2024), but also hold symbolic significance as future leaders who will influence subsequent social decision-making processes (Boca and Saraçlı, 2019). Moreover, university students tend to express values that are strongly oriented toward experience and emotion, suggesting that they may exhibit distinctive characteristics, compared with previous generations, in the ways they interpret the environment and nature (Min, 2024; Kang and Kim, 2022). A recent survey examining university students’ perceptions of forest values further indicated that their views are shaped largely by lived and emotional experiences (National Institute of Forest Science, 2023). Accordingly, investigating their environmental and spatial perceptions possesses academic significance as an important indicator for predicting the sustainability of future society (Adamczyk and Adamczyk-Kowalczuk, 2022; Leko Šimić et al, 2022). Against this backdrop, this study collected unstructured textual data from university students to derive the perceptual structure of biophilic values associated with forests through text mining-based network analysis. By examining the centrality and connectivity among these values, the study explored their hierarchical structure based on underlying tendencies. Ultimately, this study aimed to elucidate how forests function as semantic systems within urban spaces by interpreting the resulting biophilic value structure through the lens of urban environmental design.

Research Methods

Integrated Theoretical Framework

To elucidate the perceptual structure of forests among university students, representing the future generation, this study establishes an integrated theoretical framework combining biophilic cities theory, biophilic design patterns and their effects, and biophilia value typologies. First, the concept of biophilic cities proposed by Beatley and Newman (2013) conceptualizes cities as environments integrated with nature and positions urban forests as core biophilic infrastructure. Within this urban context, forests are defined not merely as green spaces, but as biophilic environments in this study.

Second, the 14 biophilic design patterns proposed by Terrapin Bright Green (2014) are classified into three categories: “Nature in the Space,” “Natural Analogues,” and “Nature of the Space.” These patterns are understood to elicit physiological and psychological effects in humans, including “Stress Reduction,” “Cognitive Performance,” and “Emotion, Mood and Preference.” In this study, these design patterns were regarded as the physical attributes of a space, while the effects associated with them were interpreted as outcomes of human cognitive and perceptual processing of those spatial attributes.

In addition, Kellert and Wilson (1993) conceptualized the human-nature relationship through biophilia values and systematized them into eight value types, including attraction, affection, reason, symbolism, and spirituality. Building upon the design patterns and their associated effects identified in previous studies, this study constructs a conceptual framework explaining how such effects are ultimately interpreted into specific value types.

In particular, this study conceptualizes forests as distinct form of biophilic infrastructure, differentiated from other types of nature experiences, and focused not merely on the interpretation of individual values but also on the relationships among them. In other words, each value type is not considered to exist independently, but rather to be structurally organized according to differences in centrality and connectivity, which together form a hierarchical value structure. Accordingly, the study examines the mechanisms behind the formation of such structures. From this perspective, the study proposes an integrated system linking urban context, environment (space), design patterns, perceived effects, value types, value structures, thereby extending biophilia hypothesis from the level of value typologies to that of value structures (see Fig. 1).

Fig. 1

Integrated theoretical framework.

Research Method

As illustrated in Fig. 2, this study employed a four-stage analytical process to investigate university students’ subjective value perceptions of forests from a biophilic perspective. In the first stage, unstructured subjective data were collected through an online survey. The participants consisted of 49 non-major undergraduate students, ranging from freshmen to seniors, enrolled at universities in the Seoul metropolitan area. Data collection was conducted from March 31 to April 6, 2023. By encouraging respondents to freely describe their holistic experiences and values related to forests without restricting their responses to predefined categories, the study sought to secure both the authenticity and depth of the data. The resulting dataset comprised subjective responses from Korean-speaking university students, offering insights into their biophilic perceptions. To enhance objectivity in data interpretation and assist international researchers and readers in grasping the semantic context, major keywords are presented in both Korean and English. In the second stage, data preprocessing was conducted. The collected textual data underwent morphological analysis using the OKT (Open Korean Text) analyzer in Python 3.8.0. To improve analytical precision, data cleaning was performed in accordance with the stopword selection principles established in previous research (Chae et al., 2023). Through this process, a total of 325 nouns, verbs, and adjectives were extracted, followed by an optimization procedure applied to the noun data. In the third stage, data analysis was performed: cluster and centrality analyses were conducted on the extracted nouns, while sentiment analysis was applied to the verbs and adjectives. In the final stage, biophilic values were hierarchically structured based on these results, and practical implications were derived by interpreting the sentiment analysis results.

Fig. 2

Research method.

Analysis Tools and Methods

The primary analytical tool used in this study was Python 3.8.0, which facilitated morphological, sentiment, and co-occurrence analyses. Based on the constructed network data, cluster analysis and network visualization were subsequently performed using Gephi 0.10.0.

For accurate preprocessing of the collected subjective responses, data cleaning was conducted in a Python 3.8.0 environment using the OKT morphological analyzer. The dataset was further refined in accordance with the stop-word selection criteria proposed by Chae et al., 2023 to improve analytical precision. Initially, nouns, verbs, and adjectives containing substantive semantic meaning were extracted, yielding a total of 325 valid words (208 nouns and 117 verbs/adjectives). A secondary optimization process was subsequently applied to improve the objectivity and replicability of the analysis. Designed to enhance data resolution, this process involved semantic optimization, whereby terms with morphological variations but equivalent meanings—such as “forest (산림)” and “in the forest (산속)”—were consolidated into representative expressions. In addition, nouns extracted from compound expressions, such as “innermost thoughts (속마음),” were incorporated into the analytical nodes. Through this systematic refinement procedure, sentiment-analysis targets (i.e., verbs and adjectives) used to assess emotional intensity were separated from the dataset, and 156 nouns representing university students’ perceptual structures toward forests were finalized as the network analysis nodes. To improve interpretive validity, these core keywords are presented concurrently in both Korean and English.

Cluster analysis was conducted to identify the semantic connectivity structure among the 156 extracted core nodes by constructing a network matrix based on co-occurrence analysis. This technique examines the organic relationships and structural associations among core keywords based on the statistical proximity of words appearing within respondents’ statements. In this study, it was applied to analyze how frequently occurring terms were organized and associated with biophilia values. To ensure objectivity in cluster derivation, a modularity-based algorithm implemented in Gephi 0.10.0 was employed, with the resolution parameter set to 1.5 to identify the optimal cluster structure. The final network consisted of 156 nodes and 1,062 edges, yielding a modularity value of 0.403. This result indicates that, under the specified analytical conditions, the network was optimally partitioned into six distinct clusters exhibiting the highest degree of structural equivalence. The derived clusters were subsequently visualized to examine inter-cluster distance and cohesion, thereby enabling interpretation of the semantic context embedded within each cluster.

Based on the cluster analysis results, a centrality analysis was conducted to elucidate the structural importance and roles of keywords within the network. This analysis quantitatively evaluates the connections among keywords, thereby identifying nodes that play a pivotal role within the network. In this study, degree centrality and betweenness centrality were estimated using Gephi 0.10.0. The resulting centrality indices were then employed in cross-cluster comparisons to distinguish clusters performing core, complementary, or peripheral roles within the overall network. Furthermore, to visually examine semantic cohesion among clusters and the statistical proximity between keywords, a co-occurrence heatmap analysis was conducted concurrently, enabling a comprehensive evaluation of both the internal cohesiveness of each cluster and its structural positioning within the network.

A sentiment analysis was performed using the KNU Korean Sentiment Dictionary (KNU SentiLex), which was developed based on the National Institute of Korean Language’s Standard Korean Language Dictionary, in order to improve analytical reliability. To effectively reflect emotional intensity within the textual data, weighted sentiment scores were calculated by multiplying each word’s sentiment polarity value by its frequency of occurrence, following the approaches proposed by Varshney and Wagh (2017) and by Elliott and Elliott (2024). Verbs and adjectives were selected as the primary targets of the sentiment analysis to examine emotional responses toward forests. The results from the sentiment analysis were ultimately used to understand how the biophilic values of forests are manifested at the affective level.

Results and Discussion

Morphological Analysis and Core Keyword Characteristics

Following morphological analysis and the removal of stopwords, a total of 325 words remained. Among these, nouns accounted for 208 words (64.0%), whereas verbs and adjectives comprised 117 words (36.0%). Since the survey specifically investigated perceptions of forests, the term “forest” itself was excluded from the analysis.

Frequency analysis revealed that the most frequently occurring nouns were “Thought” (생각; 28 occurrences, 13.46%), “Space” (공간; 13, 6.25%), “Emotion” (마음; 11, 5.29%), and “Air” (공기; 8, 3.85%). Among verbs and adjectives, the most frequent terms were “Do” (하다; 41, 35.04%), “Exist” (있다; 25, 21.37%), “Become” (되 다; 18, 15.38%), “See” (보다; 12, 10.26%), and “Feel” (8, 6.84%), as shown in Table 1. The results of the frequency analysis are also visualized in a word cloud presented in Fig. 3.

Top 10 keywords by part of speech

Fig. 3

Word cloud analysis results by part of speech.

Relatively speaking, emotionally oriented keywords such as “Thought (생각),” “Emotion (마음),” “Experience (경험),” and “Life (삶)” appeared more frequently than physically oriented terms such as “Nature (자연),” “Tree (나무),” “Animal (동물),” and “Plant (식물).” This finding suggests that respondents primarily attribute value to forests through emotional dimensions rather than through their physical elements. Furthermore, the presence of verbs and adjectives such as “See (보다),” “Scent (맡다),” and “Feel (느끼다)” indicates that forests are perceived as sensory and experiential spaces in which the five senses are actively engaged.

Taken together, these findings imply that university students perceive forests not only as emotional spaces but also as environments that facilitate experiential engagement through sensory perception.

Cluster Analysis

Cluster Analysis Results

A cluster analysis was conducted based on keyword co-occurrence frequencies, using the modularity algorithm in Gephi 0.10.0, resulting in six distinct clusters. The proportion of each cluster within the overall network and its constituent keywords are illustrated in Fig. 4.

Fig. 4

Gephi cluster analysis and network visualization.

The cluster analysis results are summarized in Table 2. Cluster 0 accounted for the largest proportion of all nodes (30.77%) and included keywords such as “Thought (생각),” “Space (공간),” “Emotion (마음),” “Tree (나무),” “Healing (치유),” “Restoration (회복),” and “Stability (안 정).” These results suggest that natural elements of forests evoke biophilic experiences corresponding to “Visual Connection with Nature” within the “Nature in the Space” category proposed by Terrapin Bright Green (2014). The emergence of the keyword “Space (공간)” also reflects aspects of the “Nature of the Space” category. Moreover, characteristics related to “Cognitive Performance” and “Emotion, Mood & Preference” were also identified within this cluster.”

Cluster analysis results: Biophilic design pattern classification and functions

Cluster 5 represented 20.51% of the network and included major keywords such as “Time (시간),” “Experience (경험),” “Nature (자연),” and “Animal (동물).” This cluster also encompassed terms such as “Family (가족),” “House (집),” “Memory (추억),” and “Sentiment (감정),” reflecting emotional attachment and a sense of belonging rooted in an affinity toward nature. These findings indicate the biophilic experiences of “Visual Connection with Nature,” “Non-Visual Connection with Nature,” and “Connection with Natural Systems” within the “Nature in the Space” category (Terrapin Bright Green, 2014), demonstrating characteristics associated with “Emotion, Mood & Preference.”

Cluster 2 also comprised 20.51% of the network, with prominent keywords including “Air (공기),” “Place (장 소),” “Smell (향기),” and “Concentration (집중).” This cluster co-occurred with terms such as “Consolation (위로),” “Innermost Thoughts (속마음),” “Tranquility (평온),” and “Purify (정화),” representing a process of connecting with the world and overcoming personal limitations through experiences in forests. These results reflect the biophilic experiences of “Thermal & Airflow Variability” and “Non-Visual Connection with Nature” within the “Nature in the Space” category (Terrapin Bright Green, 2014), exhibiting characteristics related to “Cognitive Performance.”

Comprising 16.03% of the network, Cluster 3 yielded top keywords in the following order: “Life (삶),” “Daily (일상),” “Relaxation (여유),” and “Worry (고민).” This cluster also encompassed words such as “Reflection (성찰),” “Contemplation (사색),” “Reset (리셋),” and “Discover (발견),” representing deep introspection concerning life and self-exploration through forests. These findings demonstrate the biophilic experiences of “Refuge” within Terrapin Bright Green’s (2014) “Nature of the Space” category, manifesting characteristics of “Cognitive Performance.”

Cluster 4 represented 8.97% of the network, with “Clearing (정리),” “Stroll (산책),” “Sound (소리),” and “Travel (여행)” emerging as the central keywords. Alongside terms such as “Headache (두통),” “White Noise (백색소음),” and “Feature (특색),” this cluster reflected the perception of forests as a setting for practical activities or a functional environment for achieving specific goals. This corresponds to the biophilic experiences of “Non-Visual Connection with Nature” within the “Nature in the Space” category (Terrapin Bright Green, 2014), indicating characteristics of “Stress Reduction.”

Cluster 1 accounted for the smallest proportion of the network (3.21%), with major keywords including “Healing (힐링),” “Light (빛),” “Music (음악),” and “Mood (기분)” listed in descending order of frequency. Together with terms such as “Greenery (푸름),” this cluster consisted of vocabulary that expressed forests abstractly through sensory imagery and artistic metaphors rather than through tangible physical entities. This reflects the biophilic experiences of “Visual Connection with Nature,” “Non-Visual Connection with Nature,” and “Dynamic & Diffuse Light” within the “Nature in the Space” category (Terrapin Bright Green, 2014), exhibiting characteristics associated with “Emotion, Mood & Preference.”

Overall, the biophilic design patterns observed in forests corresponded to the “Nature in the Space” and “Nature of the Space” categories proposed by Terrapin Bright Green (2014). The primary functions identified were “Cognitive Performance,” “Emotion, Mood & Preference,” and “Stress Reduction.”

To further visualize the strength of semantic cohesion both within and between clusters, a co-occurrence heatmap was generated (see Fig. 5). Beyond simple frequency counts, the heatmap serves as a core indicator of how organically specific values are interconnected within respondents’ perceptual structures. The heatmap analysis revealed that Cluster 0 exhibited the highest internal edge count (340), indicating exceptionally strong internal cohesion. Clusters 2 (115), 3 (103), and 5 (128) also demonstrated substantial cohesiveness. These high cohesion values reflect the strong contextual tendency of respondents to articulate these keywords simultaneously within individual responses. In contrast, Cluster 1 (6) and Cluster 4 (37) exhibited relatively low cohesion values, confirming their more peripheral positions within the overall network.

Fig. 5

Heatmap of connection frequencies between clusters.

Establishing a Value Structural Hierarchy Based on Centrality Analysis

To clarify the structural hierarchy of the inter-cluster network, degree centrality and betweenness centrality were analyzed. By integrating the statistical properties and semantic characteristics of each cluster, the results were mapped onto the biophilia hypothesis proposed by Kellert and Wilson (1993). This mapping revealed that perceptions of forests form a hierarchical structure composed of core, complementary, and peripheral values (See Table 3, Fig. 6).

Biophilic value hierarchy analysis

Fig. 6

Biophilic value system.

The core values were represented by Cluster 0, which played a pivotal role in driving network flow. Cluster 0 recorded the highest values across the entire network, with an average degree centrality of 0.121 and a maximum betweenness centrality of 0.176. These findings indicate that the cluster was connected to the largest number of nodes and simultaneously functioned as a critical axis mediating information flow between clusters. Semantically, Cluster 0 involves restoration and healing derived from psychological stability through appreciation of nature, corresponding to a biophilic value in which “Attraction” and “Reason” are integrated.

The complementary values were identified in Clusters 5, 2, and 3. Clusters 2 and 3 performed mutually reinforcing roles while occupying substantial proportions of the network. Specifically, Cluster 2 exhibited a relatively high maximum betweenness centrality (0.101), functioning as an intermediary that connected the network’s core and periphery through emotions related to consolation and tranquility. Notably, Cluster 2 represents a nature-interactive orientation, conceptualizing forests as spaces for sensory stimulation and relationship building. This aligns Timothy Beatley’s (2011) argument that contact with nature enhances sensory satisfaction and community solidarity. In contrast, Cluster 3 demonstrated relatively low betweenness centrality but a high average degree centrality (0.081), performing a cohesive role that internally deepened meaning through reflection and contemplation. Meanwhile, Cluster 5 corresponded to “Affection,” reflecting love and trust toward nature. Clusters 2 and 3 embodied “Spirituality,” characterized by connection with a broader world through nature that transcends the self. The combined node proportion of these two clusters reached 36.7% of the total network, constituting the largest share within a single value category. Accordingly, the complementary values correspond to the biophilia value elements of “Affection” and “Spirituality,” suggesting that “Spirituality” plays an equally vital role as the core values.

The peripheral values were represented by Clusters 4 and 1, which occupied marginal positions within the network due to their relatively low degree and betweenness centrality values. Cluster 4 was associated with activity-oriented utilitarian values such as strolling and sound, whereas Cluster 1 reflected symbolic values consumed through sensory imagery such as light and music. Although these peripheral value clusters demonstrated weaker cohesion than the core and complementary value clusters, they nevertheless embodied the biophilia value elements of “Utilitarian” and “Symbolism,” concretizing the experiential utilization of forests.

Sentiment Analysis

Analysis of the open-ended responses concerning forests revealed that neutral sentiment accounted for the largest proportion based on simple word frequency (71.8%), whereas positive sentiment (16.2%) and negative sentiment (12.0%) appeared relatively limited. However, when “weighted sentiment scores” incorporating emotional intensity were applied, positive emotion (57.9%) substantially exceeded negative emotion (42.1%). These findings confirm that perceptions of forests are fundamentally grounded in positive emotional experiences.

As shown in Table 4, positive sentiment keywords with high weighted scores included “Abundant” (많다; 10 points), “Diverse” (다양하다; 6 points), “Good” (좋다; 6 points), “Relaxed” (편안하다; 4 points), and “Beautiful” (아름답다; 4 points). These terms indicate that forests are perceived as spaces that provide psychological stability and emotional satisfaction. In particular, “Diverse” suggests that forests are recognized as environments rich in ecological components and experiential opportunities. Similarly, “Abundant” extends beyond a purely quantitative meaning and implies the richness of natural elements together with the expansion of thoughts and emotions.

Sentiment analysis

In contrast, negative sentiment keywords included “Complex” (복잡하다; −5 points), “Get away” (벗어나다; −4 points), “Fade” (지다; −3 points), and “Painful” (아프 다; −4 points). However, within their textual contexts, these expressions did not represent negative evaluations of forests themselves. Rather, they appeared to reflect a desire to escape the complexity, fatigue, and stress of everyday life. This suggests that forests function as transitional spaces that alleviate negative emotions, particularly urban stress.

Overall, the weighted sentiment score analysis demonstrates that university students perceive forests not merely as physical environments but as spaces that facilitate deep positive emotional experiences. The emergence of keywords associated with “Stress Reduction,” “Cognitive Performance,” and “Emotion, Mood & Preference” (Terrapin Bright Green, 2014) further validates the role of forests as essential biophilic infrastructure.

Conclusion

This study aimed to investigate university students’ subjective perceptions of forest values from a biophilic perspective, thereby interpreting the design patterns, functions, and value types of forests as biophilic infrastructure and deriving implications through the identification of their structures and hierarchies.

First, keywords related to forest values were extracted and categorized into nouns, verbs and adjectives. These keywords corresponded to the “Nature in the Space” (Visual Connection with Nature, Non-Visual Connection with Nature, Thermal & Airflow Variability, and Connection with Natural Systems) and “Nature of the Space” (Refuge) categories proposed by Terrapin Bright Green (2014). Accordingly, experiences derived from nature itself and from spatial characteristcs were identified as key biophilic infrastructure elements of forests. In terms of functional attributes, all three categories proposed by Terrapin Bright Green (2014)—”Stress Reduction,” “Cognitive Performance,” and “Emotion, Mood & Preference”—were represented. Among the three functions, negative expressions were primarily associated with “Stress Reduction,” whereas positive expressions were closely related to “Cognitive Performance” and “Emotion, Mood & Preference.” In addition, the findings revealed the presence of “Being Away,” one of the four characteristics of restorative environments described in Attention Restoration Theory by Kaplan and Kaplan (1989).

Second, cluster analysis revealed diverse subjective perceptions of forest values, resulting in six biophilic value types: “Attraction, Reason,” “Affection,” “Spirituality,” “Utilitarian,” and “Symbolism.” The derived clusters formed a hierarchical structure consisting of core, complementary, and peripheral values. Specifically, “Attraction” and “Reason” constituted the core values; “Affection” and “Spirituality” the complementary values; and “Utilitarian” and “Symbolism” the peripheral values. Although the core values of Attraction and Reason played central roles, the complementary value of “Spirituality” played a substantive role comparable to that of the core values. Consequently, Attraction, Reason, and Spirituality emerged as the primary biophilic value types associated with forests.

Third, sentiment analysis revealed a high proportion of neutral expressions, indicating that forests are perceived and utilized as part of everyday environments. Weighted sentiment analysis further demonstrated a predominance of positive sentiments, with terms such as “Diverse (다양 하다),” “Abundant (많다),” And “Good (좋다)” appearing most frequently. By contrast, negative expressions included “Complex (복잡하다)” and “Get away (벗어나 다).” However, rather than reflecting negative perceptions of forests, these expressions were interpreted as representing experiences associated with escaping daily life and achieving emotional restoration. These findings suggest that university students perceive the value of forests primarily through personal dimensions related to emotional and psychological recovery, rather than through public, environmental, or economic perspectives. This tendency may reflect the experiential and emotion-oriented value systems characteristic of contemporary university students, implying that different age groups may exhibit distinct structures of value perception. Furthermore, previous findings by NIFoS (2023) indicating that nature experiences significantly influence college students’ value judgments suggest that direct experiences with nature may strengthen personal perceptions of forest value. Accordingly, strategies aimed at enhancing opportunities for personal engagement with forests are needed to reinforce these value functions.

Taken together, forests as biophilic infrastructure align with the “Nature in the Space” category within the three biophilic design patterns proposed by Terrapin Bright Green (2014), while also exhibiting characteristics distinct from those of other natural environments. Moreover, university students tended to perceive forest values primarily through a personal-value-centered form of biophilia rather than through broader social values such as climate adaptation or carbon sequestration. Notably, among the eight value types proposed by Stephen Kellert and Edward O. Wilson (1993), the categories of “Aversion” and “Control” did not emerge.

These findings indicate that perceptions of forest values extend beyond simple environmental functions and reflect generational differences in the formation of experiences and values. In particular, given that climate change and urban environmental changes exert long-term influences on the lives of future generations, the value structures through which university students perceive forests hold important implications from the perspective of intergenerational justice. This suggests that how nature-based spaces are experienced and interpreted can influence the environmental awareness and behaviors of future generations, ultimately molding their perceptions of sustainability.

Based on these findings, the following implications can be drawn. The biophilic value structure identified in this study demonstrates that forests are perceived not merely as natural elements, but as environments in which sensory and cognitive experiences are integrally combined. In particular, through their correspondence with biophilic design patterns, forests can be understood as environments where Visual/Non-Visual Connections with Nature, Connections with Natural Systems, and various spatial experience elements interact in complex ways. These results suggest that when creating or improving forests in future urban environments, it is necessary to move beyond the simple provision of green spaces and adopt approaches that incorporate users’ experiential and perceptual structures. For instance, the core values “Attraction” and “Reason” can be reinforced in spaces that combine diverse sensory stimuli with exploratory experiences of nature, demonstrating the importance of multilayered experiential elements such as sound, touch, and spatial transitions, alongside visual landscapes. Furthermore, the complementary values of “Spirituality” and “Affection” indicate that forests can function as spaces that foster personal meaning-making and emotional connection, which may be associated with spatial characteristics such as “Prospect,” “Refuge,” and the experience of movement. Meanwhile, the peripheral values of “Utilitarian” and “Symbolism” suggest that forests are potential spaces for generating social meaning alongside functional use, implying that the role of forests within urban environments may expand in the long run. Overall, the findings demonstrate that forests operate as experiential environments in which multiple biophilic values are hierarchically organized, and that this value structure may serve as a foundational framework for understanding and interpreting nature-based spaces in future urban contexts.

By examining forests as biophilic infrastructure, this study establishes both a theoretical and practical foundation for understanding forests as structural environments in which personal experiences and values are formed. However, since the study focused exclusively on university students in their twenties, the findings may not be fully generalizable across other generations or cultural and linguistic groups. Nevertheless, the study is meaningful in that it identifies value perception characteristics grounded in specific generational and cultural experiences from the perspective of intergenerational justice. In addition, because the analysis focused specifically on urban forests, comparative research involving other forms of urban nature—such as parks and community gardens—is required. Future studies should further examine the universality and contextuality of the biophilic value hierarchy through cross-generational and cross-environmental comparisons. Despite these limitations, this study is significant in that it empirically operationalizes biophilia as a structured system of values rather than as an abstract concept, while also demonstrating that urban forests function as “experiential spaces” where personal meaning-making and emotional restoration occur.

References

Adamczyk J., Adamczyk-Kowalczuk M.. 2022;What do they feel, do, and expect? The young generation’s perception of environmental problems and Sustainable Development Goals in the context of quality of life. Sustainability 14(23):15551. https://doi.org/10.3390/su142315551.
Beatley T.. 2011. Biophilic cities: Integrating nature into urban design and planning Washington, DC: Island Press.
Beatley T., Newman P.. 2013;Biophilic cities are sustainable, resilient cities. Sustainability 5(8):3328–3345.
Boca G.D., Saraçlı S.. 2019;Environmental education and student’s perception, for sustainability. Sustainability 11(6):1553. https://doi.org/10.3390/su11061553.
Brown Weiss E. 1989. In fairness to future generations: international law, common patrimony, and intergenerational equity Upper Saddle River, NJ: Transnational Publishers.
Chae J.H.. 2016. The change of pluralistic values in Mt. Gwanak as a suburban mountain. Doctoral dissertation Seoul National University; Seoul, Korea:
Chae J.H., Kim S.H., Choi N., Kim T.K.. 2025;Analysis of the relationships among the value, benefit, and activities of forest culture in Korea: An application of means-chain value theory. Forests 16(2):213. https://doi.org/10.3390/f16020213.
Chae J.H., Park J.M., Kim S.H.. 2023;Restorative Environment Characteristics of an Urban Forest Based on Big Data Analytics. Forests 14(9):1770. https://doi.org/10.3390/f14091770.
Elliott J., Elliott R.. 2024. Relevance-weighted sentiment analysis of Reddit posts. In : Proceedings of the 9th Social Media Mining for Health Applications (#SMM4H’24) Workshop. p. 98–100.
Hung S.H., Chang C.Y.. 2022;How do humans value urban nature? Developing the perceived biophilic design scale (PBDs) for preference and emotion. Urban Forestry and Urban Greening 76:127730. https://doi.org/10.1016/j.ufug.2022.127730.
Jaszczak A., Kristianova K., Wasilewska O., Dunisijevic-Bojovic D.. 2020;Concepts of “Biophilia” and “Livability” in the context of social perception of public space in cities. Przestrzeń i Forma 42:133–146.
Kang Y.J., Kim S.H.. 2022;A study on consumer perception of ESG management of corporation: Focusing on the MZ generation. The Korean Journal of Advertising 33(3):7–39.
Kaplan R., Kaplan S.. 1989. The experience of nature: A psychological perspective Cambridge, UK: Cambridge University Press.
Kellert S.R., Wilson E.O.. 1993. The biophilia hypothesis Island Press. Washington, DC:
Lee Y.S.. 2023;A study on the sustainable city where nature coexists. Asia-Pacific Journal of Convergence Research Interchange 9(10):623–632.
Leko Šimić M., Sharma E., Kadlec Ž. 2022;Students’ perceptions and attitudes toward university social responsibility: Comparison between India and Croatia. Sustainability 14(21):13763. https://doi.org/10.3390/su142113763.
Min B.K.. 2024;Characteristics of social awareness and values of future generations (MZ generation): A comparative analysis of generations. Future Brief 24(6):1–4.
Mintzer I., Michel D.. 2001;Climate change, rights of future generations and intergenerational equity: an in-expert exploration of a dark and cloudy path. International Journal of Global Environmental Issues 1(2):203–222.
National Institute of Forest Science. 2023. Forest perception and emotional values across generations in Korea Seoul, Korea: Author.
Seddon N., Chausson A., Berry P., Girardin C.A.J., Smith A., Turner B.. 2020;Understanding the value and limits of nature-based solutions to climate change and other global challenges. Global Change Biology 26(3):1518–1546. https://doi.org/10.1098/rstb.2019.0120.
Söderlund J., Newman P.. 2015;Biophilic architecture: a review of the rationale and outcomes. AIMS Environmental Science 2(4):950–969.
Soga M., Gaston K.J., Yamaura Y., Kurisu K., Hanaki K.. 2016;Both direct and vicarious experiences of nature affect children’s willingness to conserve biodiversity. International Journal of Environmental Research and Public Health 13(6):529.
Terrapin Bright Green. 2014. 14 Patterns of biophilic design: improving health and well-being in the built environment New York, NY: Terrapin Bright Green, LLC.
Varshney V., Wagh R.S.. 2017;Weighted sentiment score formulation using sentence level sentiment density for opinion analysis. International Journal of Computational Intelligence Research 13(2):285–298.
Weston B.H.. 2008;Climate change and intergenerational justice: Foundational reflections. Vermont Journal of Environmental Law 9:375–430.
Wilson E.O.. 1986. Biophilia Cambridge, MA: Harvard University Press.
Yang S.Y., Lee T.K.. 2023;Analyzing research cases for indirect natural visual and auditory stimulation to measure biophilic effects. Journal of the Architectural Institute of Korea 39(10):39–49.

Article information Continued

Fig. 1

Integrated theoretical framework.

Fig. 2

Research method.

Fig. 3

Word cloud analysis results by part of speech.

Fig. 4

Gephi cluster analysis and network visualization.

Fig. 5

Heatmap of connection frequencies between clusters.

Fig. 6

Biophilic value system.

Table 1

Top 10 keywords by part of speech

Order Noun Verb/Adjective


Word Freq. % Word Freq. %
1 Thought(생각) 28 13.46 Do(하다) 41 35.04
2 Space(공간) 13 6.25 Exist(있다) 25 21.37
3 Emotion(마음) 11 5.29 Become(되다) 18 15.38
4 Air(공기) 8 3.85 See(보다) 12 10.26
5 Tree(나무) 7 3.37 Feel(느끼다) 8 6.84
6 Time(시간) 7 3.37 Scent(맡다) 7 5.98
7 Nature(자연) 6 2.88 Provide(해주다) 7 5.98
8 Experience(경험) 6 2.88 Diverse(다양하다) 6 5.13
9 Lfe(삶) 6 2.88 Live(살다) 5 4.27
10 Smell(냄새) 6 2.88 Reflect(돌아보다) 5 4.27

Table 2

Cluster analysis results: Biophilic design pattern classification and functions

Cluster % Keywords Biophilic
Design patterns Functions
Cluster 0 30.77 Thought(생각), Space(공간), Ecosystem(생태계), Emotion(마음), Tree(나무), Therapeutic(치유), Restoration(회복), Stability(안정), and others(등) Nature in the Space Cognitive Performance
Emotion, Mood & Preference
Cluster 5 20.51 Time(시간), Experience(경험), Nature(자연), Animal(동물), Plant(식물), and others(등) Nature in the Space Emotion, Mood & Preference
Cluster 2 20.51 Air(공기), Place(장소), Smell(향기), Concentration(집중), World(세상), Consolation(위로), Friend(친구), Society(사회), and others(등) Nature in the Space Cognitive Performance
Cluster 3 16.03 Life(삶), Daily(일상), Relaxation(여유), Worry(고민), Surrounding(주변), and others Nature of the Space Cognitive Performance
Cluster 4 8.97 Clearing(정리), Stroll(산책), Sound(소리), Travel(여행), Headache(두통), White noise(백색소음), Leaf(나뭇잎), and others(등) Nature in the Space Stress Reduction
Cluster 1 3.21 Healing(힐링), Light(빛), Music(음악), Mood(기분), Greenery(푸름) Nature in the Space Emotion, Mood & Preference

Table 3

Biophilic value hierarchy analysis

Hierarchy Cluster nodes Avg. Degree Centrality Max. Degree Centrality Avg. Betweenness Centrality Max. Betweenness Centrality Biophilic Values
Core 0 48 0.121 0.535 0.014 0.176 Attraction, Reason
Complementary 5 32 0.075 0.229 0.007 0.061 Affection
2 32 0.066 0.312 0.006 0.101 Spirituality
3 25 0.081 0.217 0.005 0.029
Peripheral 4 14 0.051 0.115 0.007 0.076 Utilitarian
1 5 0.028 0.064 0.002 0.011 Symbolism

Table 4

Sentiment analysis

Category Word Freq. Score Weighted Score Sum %
Positive Diverse (다양하다) 6 1 6 33 57.89%
Abundant (많다) 5 2 10
Good (좋다) 3 2 6
Get (받다) 3 1 3
Relaxed (편안하다) 2 2 4
Beautiful (아름답다) 2 2 4

Negative Complex (복잡하다) 5 −1 −5 −24 42.11%
Get away (벗어나다) 4 −1 −4
Fade (지다) 3 −1 −3
Not (아니다) 3 −2 −6
Painful (아프다) 2 −2 −4
None (없다) 2 −1 −2

Total 57 100.00%