Fair Tourism Trends and Online Discourse in the Post-Pandemic Transition: A Semantic Network Analysis

Abstract

This original empirical study examines how online discourse on fair tourism in Korea was structured and transformed during the post-pandemic tourism transition. Using unstructured online text data collected from Korean digital platforms, the study compares two periods: the pandemic continuation and early tourism recovery period (1 June 2020–31 May 2023) and the post-endemic tourism restructuring period (1 June 2023–31 May 2026). Fair tourism is conceptualized as a practical and policy-oriented discourse that connects local participation, benefit distribution, market fairness, destination governance, and tourism justice, rather than as a simple synonym for sustainable or responsible tourism. After text cleaning, morphological analysis, and keyword refinement, the top 50 core keywords for each period were analyzed using frequency analysis, degree and closeness centrality analysis, semantic network visualization, and CONCOR analysis. The results show that first-period discourse centered on regions, public policy, support projects, fair ecotourism, the social economy, resident participation, and local recovery. In contrast, second-period discourse was more strongly associated with foreign tourists, international tourism recovery, accommodation and service use, price fairness, overcharging and unfairness controversies, and sustainable destination governance. This study extends fair tourism research by revealing how fair tourism is constructed and rearticulated through digitally mediated public discourse.

1. Introduction

The COVID-19 pandemic fundamentally disrupted the growth trajectory and value structure of the contemporary tourism industry. Debates on tourism recovery have therefore moved beyond the simple restoration of demand and have increasingly centered on structural transformation involving sustainability, resilience, community-centered values, and digital transition (Gössling et al., 2020; Sharma et al., 2021; Sigala, 2020). Before the pandemic, the tourism industry had largely developed within a quantitative growth paradigm driven by the expansion of international tourist arrivals, tourism consumption, and destination development. However, the sudden contraction of international tourism demand, mobility restrictions, social distancing measures, and the decline of local tourism economies after COVID-19 revealed that tourism is not only vulnerable to external shocks but also deeply embedded in local economic and social systems (Brouder, 2020; Higgins-Desbiolles, 2020; Niewiadomski, 2020). During the prolonged pandemic, tourists also began to show greater interest in nature-based travel, small-scale tourism, short-distance and domestic trips, and community-based travel rather than crowded forms of mass tourism (Romagosa, 2020; Sung et al., 2021). These changes intensified the need to rethink tourism recovery around sustainability, fairness, and local coexistence rather than a simple return to pre-pandemic growth patterns (Ioannides & Gyimóthy, 2020).

As tourism demand has recovered rapidly in the post-pandemic period, major destinations have also experienced renewed pressures. International tourism was reported to have recovered to pre-pandemic levels in 2024, and this recovery has brought renewed attention to issues such as visitor concentration in popular destinations, traffic congestion, waste generation, declining quality of life among local residents, and the touristification of residential and commercial spaces (UN Tourism, 2025; Koens et al., 2018; Dodds & Butler, 2019; Cocola-Gant et al., 2020). Although tourism recovery can contribute to local economic revitalization and the normalization of tourism businesses, it also raises important questions about how the benefits and costs generated by tourism development and consumption are distributed among stakeholders. In this context, fair tourism needs to be reconsidered as an important alternative discourse in the post-pandemic tourism transition (Rastegar et al., 2021; Rastegar & Ruhanen, 2022; Jamal & Camargo, 2018).

Fair tourism refers to a form of tourism that seeks to ensure a fairer distribution of benefits and costs among tourists, local residents, tourism businesses, communities, natural environments, and cultural resources, while emphasizing local participation and respect for culture and the environment (Goodwin & Roe, 2001). In this study, fair tourism is understood as a practical and policy-oriented discourse that connects sustainability, responsibility, market fairness, local participation, and tourism justice in the Korean online and policy context. A more detailed conceptual positioning of fair tourism in relation to adjacent tourism concepts is provided in the following section.

In international tourism scholarship, discussions similar to fair tourism have developed more frequently through concepts such as responsible tourism, sustainable tourism, ethical tourism, pro-poor tourism, community-based tourism, and tourism justice rather than through the term fair tourism itself (Caruana et al., 2014; Scheyvens, 1999; Jamal & Camargo, 2018). Research on responsible tourism has emphasized that tourists should consider the environmental and sociocultural consequences of their travel and act in ways that enhance positive impacts on host communities (Caruana et al., 2014). In parallel, tourism justice scholarship has drawn attention to questions of who benefits from tourism development, who bears its costs, and how tourism-related harms can be recognized and restored (Jamal & Camargo, 2018; Jamal & Higham, 2021; Rastegar & Ruhanen, 2022). These debates provide an important theoretical foundation for understanding fair tourism. However, they also suggest that fair tourism should not be reduced to individual tourist ethics or alternative tourism products. Instead, it should be examined as a broader discursive field in which sustainability, responsibility, market fairness, local participation, and tourism justice are connected and reinterpreted in specific social and policy contexts.

In Korea, fair tourism has also been the subject of sustained academic attention. Early studies focused on clarifying the concept and components of fair tourism, while subsequent research expanded to fair tourism perception, ethical consumption, tourism attitudes, behavioral intention, place attachment, responsible tourism behavior, perceptions of fair tourism destinations, willingness to pay a premium, and tourism ESG (Hwang & Lee, 2011; Yeo et al., 2012; T.-S. Lee & Kim, 2015, 2018; K.-P. Lee & Kwon, 2016; Kang, 2024). These studies have contributed to explaining how individuals perceive fair tourism and how such perceptions are linked to attitudes and behavioral intentions. Nevertheless, survey-, interview-, and case-based approaches have limitations in capturing how fair tourism is socially constructed, circulated, and reinterpreted in online public discourse. In particular, they cannot fully explain which keywords organize fair tourism discourse, how those keywords are semantically connected, and how the discourse changes across different stages of post-pandemic tourism recovery. This study addresses this gap by shifting the analytical focus from individual-level attitudes and behavioral intentions to the relational structure of online discourse.

Online spaces function as digitally mediated discursive arenas in which tourists, local residents, tourism businesses, public institutions, and other stakeholders produce and share perceptions, experiences, complaints, policy information, and value-based claims related to tourism. Unstructured text generated through blogs, online communities, web documents, and other digital platforms can therefore serve as useful data for identifying public attention, perception structures, issue dynamics, and discursive connections surrounding a specific tourism phenomenon (Xiang & Gretzel, 2010; Zeng & Gerritsen, 2014; Drieger, 2013). In this respect, online discourse analysis is closely connected to the digital transformation of tourism research because it enables researchers to examine tourism meanings as they are produced and circulated through digital platforms. Semantic network analysis is a useful method for moving beyond simple word frequency by examining the relationships among keywords extracted from unstructured text (Drieger, 2013). Centrality analysis can identify structurally important keywords within an online discourse, while CONCOR analysis can classify keywords with similar connection patterns and reveal the substructures of discourse clusters (Freeman, 1979; Wasserman & Faust, 1994).

Against this research background, this original empirical study applies semantic network analysis to original online unstructured text data in order to compare the structure and temporal changes of online discourse on fair tourism during the post-pandemic transition. It does not aim to synthesize prior literature as a review article; rather, it empirically examines how fair tourism discourse was constructed and transformed through keyword relations and semantic clusters in Korean digital platforms. Specifically, the study first identifies differences in the core keywords related to fair tourism between the pandemic continuation and early tourism recovery period and the post-endemic tourism restructuring period. Second, it examines changes in the semantic connections and central keywords between the two periods in order to understand the structural characteristics of fair tourism discourse. Third, it uses CONCOR analysis to derive sub-clusters of fair tourism discourse and interpret the discursive meaning of each cluster by period. Based on these findings, the study discusses how online fair tourism discourse has become connected with sustainable tourism, responsible tourism, local coexistence, ESG, market fairness, and tourism justice in the post-pandemic tourism context. By doing so, this study contributes to fair tourism research by demonstrating how social big data and semantic network analysis can reveal the changing meanings of fair tourism as a digitally mediated public discourse.

2. Literature Review

2.1. Conceptual Positioning of Fair Tourism and Related Tourism Discourses

Fair tourism is grounded in the idea that tourists should respect the lives and cultures of local residents, that economic benefits generated through tourism consumption should be returned to local communities, and that tourism activities should contribute to the conservation of natural environments and cultural heritage (Hwang & Lee, 2011; Goodwin & Roe, 2001). Building on this foundation, this study positions fair tourism as a relational and policy-oriented discourse that foregrounds how tourism benefits, costs, responsibilities, and decision-making opportunities are distributed among tourists, local residents, tourism businesses, communities, and public institutions.

Sustainable tourism provides a broad development principle that balances the current and future economic, sociocultural, and environmental impacts of tourism in order to meet the needs of visitors, the tourism industry, the environment, and host communities, whereas responsible tourism places greater emphasis on practical actions through which tourists, tourism businesses, local communities, and policy actors reduce negative impacts and increase positive outcomes in tourism settings (Bramwell & Lane, 1993; Mihalic, 2016). Ethical tourism foregrounds tourists’ moral choices and consumption practices, ecotourism emphasizes nature-based conservation and environmental learning, and community-based tourism focuses on local participation and community benefits (Caruana et al., 2014; Scheyvens, 1999; Choi & Sirakaya, 2006). Fair tourism overlaps with these concepts, but its distinctive contribution lies in foregrounding fairness as a cross-cutting issue that connects ethical consumption, community participation, benefit distribution, market order, policy accountability, and destination governance. From this perspective, fair tourism is not treated as an all-encompassing umbrella concept, but as a specific discursive category through which broader sustainability and justice debates are articulated in tourism contexts. Recent international research on tourism fairness further supports this relational positioning. Banerjee et al. (2023) conceptualized fairness in tourism as a multistakeholder issue involving tourists, tourism service providers, digital platforms, local communities, and environmental interests. This perspective suggests that fairness in tourism should be evaluated not only from the viewpoint of individual tourists but also in terms of how benefits, visibility, opportunities, responsibilities, and tourism-related burdens are distributed across multiple stakeholders.

In Korea, fair tourism research has gradually expanded from early discussions of tourist behavior and destination management to more recent debates on tourism ESG. Hwang and Lee (2011) conceptualized fair tourism as a matter of fair relationships and benefit distribution among tourists, local residents, the ecological environment, and the tourism industry, emphasizing fair tourism as a practical form of sustainable tourism. Subsequent studies have moved beyond the normative definition of fair tourism and have examined fair tourism behavior based on the theory of planned behavior, fair tourism perception and purchase intention, willingness to pay a premium, perceptions and attitudes toward fair tourism destinations, and the typology of tourism ESG (T.-S. Lee & Kim, 2015, 2018; K.-P. Lee & Kwon, 2016; Kang, 2024). These studies have contributed to clarifying the behavioral, perceptual, and managerial dimensions of fair tourism in the Korean context. In Korea, fair tourism has developed not only as an ethical travel practice but also as a policy and public discourse category associated with regional tourism, local coexistence, social economy, resident participation, and sustainable destination management.

Nevertheless, previous fair tourism research has tended to focus on individual tourists’ attitudes, behavioral intentions, willingness to pay, or evaluations of specific destinations and cases. Such approaches are useful for explaining how tourists perceive fair tourism and how those perceptions influence behavioral outcomes. However, they provide limited insight into how fair tourism is socially constructed, circulated, and reinterpreted in online public discourse. In particular, survey-, interview-, case-, and document-based studies cannot fully explain which keywords organize fair tourism discourse, how those keywords are semantically connected, and how the central meanings of fair tourism change during different stages of tourism recovery. This limitation is important because the meaning of fair tourism is not fixed; it can be reconstructed according to changing social conditions, policy agendas, tourism market issues, and stakeholder concerns. Therefore, the present study approaches fair tourism not merely as an individual-level attitude or behavioral intention, but as a socially constructed discourse structure embedded in online public communication.

More recently, international tourism scholarship has increasingly extended discussions related to fair tourism toward the perspective of tourism justice. Tourism justice raises questions about who can participate in tourism policy and destination management decision-making and how local cultures and identities are recognized in the process of tourism commodification (Jamal & Higham, 2021; Rastegar & Ruhanen, 2022). This perspective expands fair tourism beyond individual ethical choice or alternative tourism products and situates it within broader issues of structural fairness, including distributive, procedural, recognitional, and restorative justice in tourism systems. Therefore, this study understands fair tourism as a practical and policy-oriented discourse connected to sustainable and responsible tourism, while also functioning as a context-specific expression of tourism justice articulated through online discourse during the post-pandemic tourism transition.

2.2. Post-Pandemic Tourism Transition and Changes in Fair Tourism Discourse

The post-pandemic tourism transition cannot be understood simply as a cyclical process of decline and recovery in tourism demand. Rather, it represents a structural reconfiguration of tourism mobility, tourist values, destination management priorities, and stakeholder relationships. During the COVID-19 pandemic, tourist mobility was strongly constrained by social distancing measures, travel restrictions, and the desire to avoid crowded places. In this context, tourists showed growing interest in short-distance travel, nature-based tourism, and safer forms of tourism rather than large-scale mobility and crowded mass tourism (Romagosa, 2020; Sung et al., 2021). These changes suggest that tourist choices can be reshaped not only by personal satisfaction or price utility, but also by value-based criteria such as safety, responsibility, and environmental concern. Therefore, the pandemic period provided a context in which fair tourism could be discussed in connection with responsible travel, safer mobility, local consumption, and community-based recovery.

However, as tourism demand recovered rapidly after the endemic transition, fair tourism discourse entered a new phase. The recovery of international tourism and the reactivation of both domestic and inbound tourism demand in Korea may contribute to the normalization of the tourism industry and the recovery of local economies. At the same time, this recovery has also brought renewed attention to overtourism-related problems, including visitor concentration in major destinations, traffic congestion, waste generation, encroachment on residents’ everyday living spaces, and tourism gentrification (UN Tourism, 2025; Koens et al., 2018; Dodds & Butler, 2019; Cocola-Gant et al., 2020). In particular, when the benefits of tourism recovery are not evenly distributed across local communities, or when environmental and sociocultural costs are concentrated among certain groups of residents, tourism recovery itself may generate new forms of unfairness. This means that the post-pandemic recovery of tourism should not be regarded merely as the restoration of demand or the normalization of the industry; rather, it requires a restructuring of tourism toward a more sustainable, resilient, inclusive, and fair system (Abbas et al., 2021).

In this context, fair tourism can be reinterpreted as a key discourse through which the fairness of tourism recovery is debated in the post-pandemic tourism transition. During the pandemic continuation and early tourism recovery period, fair tourism may have been associated with responsible travel, domestic tourism, nature-based tourism, and local consumption. By contrast, during the post-endemic tourism restructuring period, fair tourism may become more closely connected to overtourism, local coexistence, tourism ESG, sustainable tourism, and tourism justice. This shift does not necessarily indicate direct changes in actual tourism practices or institutional governance. Rather, it suggests that the meanings attached to fair tourism in online discourse may change according to broader social conditions, tourism market recovery, and emerging destination issues. Accordingly, fair tourism discourse can expand beyond individual tourists’ ethical consumption or responsible travel behavior and develop into a structural discourse that addresses the distribution of benefits and costs in tourism development and consumption, residents’ participation, fairness in destination management, and justice in tourism recovery (Rastegar et al., 2021).

Therefore, comparing fair tourism discourse between the pandemic continuation period and the post-endemic transition is not merely a matter of identifying differences in keywords across periods. Rather, it is a process of examining how fair tourism was discursively constructed during a period of crisis and how it was later rearticulated during the recovery and restructuring of the tourism market. This study does not assume that online discourse directly represents actual changes in tourism behavior or governance. Instead, it examines how online discourse reflects changing public attention, issue framing, and semantic associations surrounding fair tourism. From this perspective, the present study compares online discourse on fair tourism across the two periods in order to explore what social meanings fair tourism has acquired as tourism has been debated in relation to sustainability, fairness, local coexistence, market order, and destination governance in the post-pandemic era.

2.3. Tourism Social Big Data and Semantic Network Analysis

Digital transformation in tourism research has changed the ways in which tourist experiences and destination perceptions are formed, shared, and circulated. Tourists actively use online platforms when searching for information before travel, recording experiences during travel, and evaluating or recommending destinations after travel. In this process, various forms of unstructured text data are accumulated through blogs, online reviews, community posts, and other digital channels. Such user-generated content can serve as an important data source for analyzing tourism phenomena because it contains tourists’ experiences, emotions, satisfaction and dissatisfaction, destination images, evaluations of tourism products, and responses to social issues (Lu & Stepchenkova, 2015; Marine-Roig & Anton Clavé, 2016). Whereas traditional surveys and interviews collect data based on variables and questions predetermined by researchers, online unstructured data allow researchers to explore naturally occurring perceptions and discourse structures through language voluntarily produced by users.

At the same time, online tourism data should not be regarded as a neutral or complete representation of all tourists, residents, or stakeholders. Platform-specific user demographics, search algorithms, content exposure mechanisms, promotional content, and public-sector communication can influence what is visible in online discourse. Therefore, social big data analysis is most useful when it is used to identify digitally mediated patterns of attention, association, and discourse formation, rather than to claim direct representativeness of actual tourism behavior. This point is particularly important for fair tourism because fair tourism is not only a matter of tourist attitudes or behavioral intentions, but also a value-laden discourse shaped by public policy, local community issues, tourism market fairness, destination governance, and stakeholder conflicts.

In recent years, tourism and hospitality studies have increasingly used various forms of big data, including social media data, online reviews, search data, and mobile data, to analyze tourism demand, destination image, tourism policy issues, and tourism trends. This line of research suggests that tourism big data are not merely large-scale datasets, but analytical resources that can capture changes in tourist and public perceptions as well as market responses in a relatively timely manner (Li et al., 2018; Mariani et al., 2018; Berezina et al., 2016). However, many studies using tourism big data have focused on destination image, satisfaction, review valence, or demand prediction, while relatively less attention has been paid to how normative and policy-related tourism concepts are discursively constructed in online public spaces. Recent international studies have also begun to examine how digital tourism systems may produce or mitigate fairness-related problems. Rahmani et al. (2022) showed that the incorporation of contextual information into point-of-interest recommendation systems can affect not only recommendation accuracy but also fairness for both users and tourism service providers. Similarly, Dela Cruz et al. (2024) demonstrated that efforts to improve consumer and provider fairness in contextual point-of-interest recommendations may involve trade-offs between recommendation accuracy and equitable exposure. Fair tourism is also a complex topic involving values, responsibility, local communities, the environment, policy, and justice. Therefore, its social meaning cannot be fully explained only through the attitudes or behavioral intentions of specific tourist groups. Analyzing online unstructured text related to fair tourism is useful for identifying the words with which fair tourism is discussed, the semantic relationships through which it is constructed, and the sub-discourses into which it is differentiated in online spaces.

Semantic network analysis is a method for identifying the meaning structure of a given phenomenon by examining not only the frequency of keywords in text, but also the relationships and structural positions among them (Drieger, 2013). Conventional frequency analysis can show which words are mentioned frequently in relation to a particular topic, but it does not sufficiently explain how those words are connected to one another or what positions they occupy within the overall discourse structure. By contrast, semantic network analysis constructs a semantic network based on keyword co-occurrence relationships and uses centrality measures, such as degree centrality and closeness centrality, to identify the structural roles of core and connecting keywords within a discourse (Freeman, 1979; Wasserman & Faust, 1994). Degree centrality indicates the extent to which a keyword is directly connected to other keywords, while closeness centrality explains how efficiently a keyword can reach other keywords within the entire network. In the context of fair tourism, this approach is useful because it can show whether fairness-related discourse is organized around ethical travel, regional recovery, public policy support, tourism market fairness, resident participation, or destination governance.

In addition, CONCOR analysis can be used to identify the substructures of online discourse by clustering keywords with similar connection patterns (Breiger et al., 1975). Compared with approaches that only list frequent keywords or interpret isolated terms, CONCOR analysis allows researchers to examine groups of keywords that occupy similar relational positions within the network. This is appropriate for the present study because fair tourism discourse is multidimensional and can be connected to sustainable tourism, responsible tourism, local coexistence, ethical tourism, tourism policy, ESG, overtourism, price fairness, and destination governance. Semantic network analysis and CONCOR analysis therefore make it possible to systematically examine what kinds of relational networks are formed among fair tourism-related keywords, which keywords occupy central positions within the network, and how different discourse clusters are distinguished from one another. Accordingly, these methods are well suited to analyzing fair tourism as a social meaning system constructed within online public discourse, while also acknowledging that the findings represent the structure of online discourse rather than direct evidence of actual tourist behavior or institutional change.

3. Research Design

3.1. Research Questions and Analytical Framework

This study adopts an exploratory and comparative empirical design based on online unstructured text data collected from Korean digital platforms to examine how online discourse on fair tourism was structured and transformed between the pandemic continuation and early tourism recovery period and the post-endemic tourism restructuring period. Previous studies on fair tourism have primarily examined tourists’ attitudes, behavioral intentions, perceptions of fair tourism, and willingness to pay a premium through surveys, interviews, and case studies. Although these approaches have contributed to understanding individual-level perceptions and behavioral responses, relatively little research has investigated which core keywords structure fair tourism discourse in online spaces, how these keywords form semantic connections, and how the discourse changes across different stages of the post-pandemic tourism transition. To address this gap, the present study applies social big data-based semantic network analysis to identify the structural characteristics and temporal changes of online discourse on fair tourism. The analytical framework consists of four complementary stages. First, frequency analysis is used to identify the most salient keywords in each period. Second, degree and closeness centrality analyses are conducted to examine the structural importance and accessibility of keywords within the semantic network. Third, semantic network visualization is used to compare the overall relational structure of fair tourism discourse between the two periods. Fourth, CONCOR analysis is applied to identify sub-discourse clusters composed of keywords with similar connection patterns. By applying the same analytical procedure to both periods, this study compares changes in the semantic structure of fair tourism discourse while acknowledging that the findings represent online discourse patterns rather than direct evidence of actual tourism behavior or institutional change.

Research Question 1. What differences and characteristics can be observed in the frequency of core keywords related to fair tourism between the pandemic continuation and early tourism recovery period and the post-endemic tourism restructuring period?

Research Question 2. How do fair tourism-related keywords differ between the two periods in terms of centrality and semantic connection structures within the semantic network?

Research Question 3. What semantic differences can be identified in the fair tourism discourse clusters derived through CONCOR analysis across the two periods?

3.2. Analysis Period and Data Collection

To analyze temporal changes in online discourse on fair tourism, this study set the data collection period as six years, from 1 June 2020 to 31 May 2026. Rather than analyzing the entire period as a single integrated network, the data were divided into two three-year periods in order to clearly compare discursive differences between the pandemic continuation period and the post-endemic tourism restructuring period. Period 1, from 1 June 2020 to 31 May 2023, represents a period in which the prolonged COVID-19 pandemic, social distancing measures, mobility restrictions, the shift toward domestic tourism, and early tourism recovery discourse coexisted. Period 2, from 1 June 2023 to 31 May 2026, represents a period after the downgrading of the COVID-19 crisis alert level in Korea, during which tourism demand recovered, international tourism resumed, overtourism re-emerged, and discussions on ESG and local coexistence in tourism expanded.

Data collection was conducted using TexTom, an online unstructured text collection and analysis tool. The exact search query used for data collection was the Korean term “공정관광,” which is translated as “fair tourism” in this manuscript. Because the related Korean expression “공정여행,” translated as “fair travel,” is sometimes used with a similar meaning in Korean online spaces, its overlap and semantic relevance were reviewed during the preliminary search and data-cleaning stages. However, the main search query was centered on “공정관광” in accordance with the research purpose and analytical scope. When expressions such as “공정여행” appeared together with fair tourism-related content, they were reviewed in context and retained only when they were semantically relevant to the broader discourse on fair tourism. The data collection channels included major Korean online platforms and web sources: Naver Blog, Naver Café, Naver Knowledge iN, Naver web documents, Daum Blog, Daum Café, Daum web documents, Google web documents, and Facebook. These sources were selected because they contain diverse forms of tourism-related online discourse, including tourists’ experiences, opinions from local residents and civil society, public-sector information, news-based web documents, and travel-related online discussions. Since the purpose of this study is to examine online discourse rather than to estimate the perceptions of all tourists or residents, these platforms were used as digital arenas in which fair tourism-related meanings, issues, and associations are produced and circulated.

Documents were included in the dataset when they directly mentioned fair tourism or discussed fair tourism-related programs, policies, destinations, tourism practices, local participation, sustainability, market fairness, or tourism justice in a relevant tourism context. Documents were excluded if they were duplicates; purely promotional or advertising materials without substantive relevance to fair tourism; simple link collections; documents unrelated to tourism; texts consisting mainly of meaningless symbols or numbers; or documents that mentioned the search term only incidentally, without any connection to tourism discourse, policy, destination management, or tourism consumption. As a result of data collection, 6968 raw documents, corresponding to 3.90 MB, were collected for Period 1, and 6747 raw documents, corresponding to 3.83 MB, were collected for Period 2. After removing duplicate documents, advertising content, documents not directly related to the search term, simple links, and meaningless characters, the cleaned text data were used to derive the top 50 core keywords for each period. To systematically analyze temporal differences in online discourse on fair tourism, the research procedure was organized into four stages: data collection, data preprocessing, keyword extraction, and semantic network analysis and visualization. The detailed research process is presented in Figure 1.

To enhance transparency regarding the composition of the online text corpus, the number of raw documents collected from each platform was recorded separately. In Period 1, 6968 raw documents were collected, including 4004 Naver Blog posts, 984 Naver Café posts, 433 Naver Knowledge iN documents, 659 Daum Blog posts, 658 Daum Café posts, and 230 Google/Facebook-related web documents. In Period 2, 6747 raw documents were collected, including 3975 Naver Blog posts, 775 Naver Café posts, 443 Naver Knowledge iN documents, 658 Daum Blog posts, 656 Daum Café posts, and 240 Google/Facebook-related web documents. To further clarify the corpus construction process,

Table 1. Corpus Construction and Cleaning Summary.

Corpus Construction Stage	Period 1: 1 June 2020–31 May 2023	Period 2: 1 June 2023–31 May 2026
Raw documents collected	6968	6747
Documents removed during cleaning	325	244
Final documents retained after cleaning	6643	6503
Core keywords retained for network analysis	50	50

summarizes the number of raw documents collected, documents removed during cleaning, and final documents retained for each analytical period. To enhance methodological transparency and reproducibility, representative examples of excluded documents, incidentally matched documents, stop words, merged synonyms, standardized compound nouns, and retained or removed foreign-language expressions are presented in Appendix A.

3.3. Data Preprocessing and Keyword Selection

The collected unstructured text data underwent a step-by-step cleaning process to improve the reliability and consistency of the analysis. First, duplicate posts, advertising content, repetitive promotional texts, documents unrelated to the search term, simple link collections, meaningless symbols and numbers, and other content unsuitable for semantic analysis were removed from the dataset. Foreign-language expressions were removed only when they were irrelevant to the interpretation of Korean fair tourism discourse; foreign words or proper nouns that carried contextual meaning, such as country names, destination names, or tourism-related terms, were retained. Next, keywords were extracted mainly from nouns through morphological analysis. Because this study aims to examine the semantic structure of online discourse, particles, conjunctions, general modifiers, and overly generic terms that did not contribute to the interpretation of fair tourism discourse were treated as stop words. Similar terms and synonyms were integrated into the same semantic units, and compound nouns were standardized to ensure consistency in keyword interpretation. For example, expressions referring to the same concept were unified where appropriate, while terms with different contextual meanings were retained as separate keywords. Compound nouns related to tourism policies, regional tourism programs, fair tourism projects, tourism markets, and destination management were reviewed in their original textual contexts before standardization. To enhance the validity of keyword refinement, two professors in tourism studies reviewed the criteria for stop-word removal, synonym integration, and compound noun processing. In particular, the reviewers discussed whether to include or exclude general and administrative terms with limited relevance to fair tourism, as well as keywords that carried contextual meaning in relation to fair tourism-related policy projects, contests, regional tourism programs, the social economy, tourism market fairness, and destination governance.

Keyword selection was conducted by considering both frequency and relevance to the research topic. In the first stage, high-frequency keywords were extracted for each period. In the second stage, their semantic relevance to fair tourism discourse was reviewed by examining whether each keyword contributed to explaining the meaning structure of fair tourism in the corresponding period. Subsequently, two tourism scholars conducted an expert review to confirm whether each keyword was appropriate for explaining the semantic structure of online discourse on fair tourism. During this review process, the keywords were classified by distinguishing between those with high frequency but limited interpretive value, those reflecting the policy, market, or regional tourism context of a specific period, and those connected to the core values of fair tourism. Based on this process, the final keyword list was confirmed. The top 50 core keywords for each period were used in the final analysis. The number of keywords was determined by considering previous social big data and semantic network analysis studies, which commonly used approximately 40 to 70 keywords for network analysis, as well as the readability and interpretability of the network structure. In this study, the top 50 keyword threshold was considered appropriate because it allowed the analysis to include the main thematic elements of fair tourism discourse while avoiding excessive network complexity caused by low-frequency or weakly interpretable keywords. The final selected keywords were used for frequency analysis, centrality analysis, semantic network visualization, and CONCOR analysis for each period (Han & An, 2022; Kim & Kim, 2023).

To ensure consistency in keyword refinement, the same cleaning criteria were applied to both periods. This procedure was adopted to reduce the possibility that differences in keyword frequency, centrality, network structure, and clusters between the two periods were caused by inconsistent preprocessing decisions. In addition, the criteria for synonym integration and stop-word removal, where subjective judgment by the researchers could be involved, were separately recorded to enhance analytical transparency. The final keyword list was confirmed through cross-review by the researchers and two professors in tourism studies. For keywords on which opinions differed, the original textual context and relevance to the research purpose were re-examined before deciding whether to include them in the final analysis.

3.4. Semantic Network and CONCOR Analysis

This study conducted semantic network analysis and CONCOR analysis to identify the semantic structure of online discourse on fair tourism. Semantic network analysis is a method that constructs a semantic network based on co-occurrence relationships among keywords extracted from text and analyzes the structural position of each keyword within the network (Doerfel, 1998; Drieger, 2013). In particular, centrality analysis is useful for identifying the structural importance of specific keywords within a network (Freeman, 1979; Wasserman & Faust, 1994). Morphological analysis was conducted using the Korean text-processing and morphological analysis functions embedded in TexTom. Keywords were extracted mainly from nouns, and the final semantic network was constructed using the top 50 keywords selected for each period. The keyword co-occurrence matrix was constructed based on the document-level co-occurrence matrix generated by TexTom. In this study, co-occurrence was defined at the document level; that is, two keywords were considered connected when they appeared together within the same collected document. No additional sentence-level or paragraph-level co-occurrence window was applied. Each cell value in the matrix indicates the frequency with which a specific pair of keywords co-appeared within the same document. Accordingly, the co-occurrence matrix was used as weighted relational data showing which keywords were discussed together within the same document context in online discourse on fair tourism. The top-50 keyword threshold served as the network inclusion criterion, and the resulting keyword co-occurrence matrices were treated as weighted networks for centrality and CONCOR analyses in UCINET and NetDraw.

UCINET was used to calculate centrality measures based on the keyword relationship matrix, while NetDraw was used to visualize the structure of the semantic network. This study focused on degree centrality and closeness centrality to analyze the core keywords of online discourse on fair tourism. Degree centrality refers to the extent to which a specific keyword is directly connected to other keywords, and it is useful for identifying core terms with strong direct associations in fair tourism discourse. Closeness centrality indicates how closely a keyword is connected to other keywords in the network through short paths, and it is used to identify keywords that are highly accessible and centrally positioned within the overall semantic network. Therefore, keywords with high degree centrality can be interpreted as core terms with strong direct connectivity, whereas keywords with high closeness centrality can be interpreted as central concepts that are efficiently connected to the overall discourse structure (Borgatti, 2005; Opsahl et al., 2010).

CONCOR, or CONvergence of iterated CORrelations, is a clustering method that groups keywords with similar connection patterns based on structural equivalence. In this study, CONCOR analysis was applied to the keyword networks for each period in order to derive the sub-clusters of online discourse on fair tourism (Song et al., 2022; Tang et al., 2024). CONCOR analysis was considered appropriate for this study because the purpose of the analysis was not simply to group keywords according to surface-level similarity or frequency, but to identify groups of keywords that occupy similar relational positions within the semantic network. Compared with clustering approaches that mainly rely on distance-based similarity, CONCOR is useful for examining structural equivalence among keywords and for identifying discourse substructures formed by similar patterns of connection. This is particularly relevant to fair tourism because fair tourism discourse is multidimensional and may be connected to public policy, regional recovery, social economy, responsible travel, overtourism, price fairness, destination governance, ESG, and tourism justice. The clusters derived through CONCOR analysis were named by considering both the semantic commonality of the keywords within each cluster and the network structure. To reduce researcher subjectivity in cluster interpretation, the initial cluster labels were first proposed by the researchers and then reviewed by two professors in tourism studies. When there were differences in interpretation, the representative keywords, co-occurrence patterns, and original text contexts were re-examined, and the final cluster names were determined through consensus. This procedure was used to ensure that the cluster labels reflected not only individual keyword meanings but also the broader discourse structure represented by each cluster. The results are presented with a focus on comparison between the two periods. First, the frequencies of the top keywords in each period were compared to identify changes in the most frequently mentioned terms of fair tourism discourse. Second, centrality analysis was used to compare core keywords with strong direct connectivity and central keywords with high accessibility within the overall semantic network. Third, NetDraw was used to visualize the semantic network for each period and to interpret the structural differences in fair tourism discourse. Fourth, the clusters derived through CONCOR analysis were compared to examine how the substructures of fair tourism discourse changed between the pandemic continuation period and the post-endemic transition. Each analytical step was therefore used for a distinct purpose: frequency analysis identified salient terms, centrality analysis identified structurally important keywords, visualization showed relational patterns, and CONCOR analysis revealed discourse clusters.

Rather than integrating the full six-year dataset into a single network, this study analyzed the two periods independently and then compared the results. This approach was adopted to more clearly reveal discursive differences between the pandemic continuation period and the post-endemic transition. An integrated analysis of the full period could show the general structure of fair tourism discourse, but it could also obscure period-specific differences in keyword centrality, network structure, and discourse clusters. Therefore, by setting two equivalent three-year analytical periods and applying the same analytical procedures to both, this study sought to examine temporal changes and structural transitions in online discourse on fair tourism in a more systematic manner. However, the results should be interpreted as changes in online discourse structures rather than as direct evidence of changes in actual tourist behavior, institutional governance, or destination management practices.

4. Results

4.1. Frequency Analysis of Fair Tourism Keywords

Table 2. Top 50 Keywords Related to Fair Tourism by Analytical Period.

Rank	Period 1: Pandemic Continuation and Early Tourism Recovery Period		Period 2: Post-Endemic Tourism Restructuring Period
	Keyword	Frequency	Keyword	Frequency
1	Region	404	Foreigners	564
2	Ministry of Culture, Sports and Tourism	373	China	343
3	China	325	Daejeon	314
4	Support	319	Program	295
5	Program	300	Ministry of Culture, Sports and Tourism	273
6	Culture	281	Culture	244
7	Development	248	Tourists	243
8	Project	242	Korea	239
9	Tourist Destination	229	Tourist Destination	193
10	Contest	221	Busan	187
11	Seoul	220	Japan	182
12	Korea	184	History	171
13	Operation	179	Hotel	170
14	World	175	United States	168
15	Gyeonggi-do	169	Development	165
16	Information	167	News	165
17	COVID-19	167	Industry	164
18	Tourists	165	Seoul	164
19	Revitalization	164	Project	156
20	Japan	164	Market	155
21	Economy	159	Experience	146
22	Jeju	156	Planner	140
23	Education	148	Event	139
24	Environment	146	Nationwide	134
25	Korea Tourism Organization	140	Education	133
26	Company	140	Contest	133
27	Hotel	135	Global	133
28	Industry	134	Service	132
29	Event	131	Unfairness	130
30	Content	130	Accommodation	128
31	United States	129	Jeju	126
32	History	126	Fairness	126
33	Fair Trade	121	Controversy	123
34	Promotion	121	Reservation	119
35	Fairness	120	Price	117
36	Fair Ecotourism	97	Tour	104
37	Gangwon-do	97	Fair Trade	100
38	Fostering	92	Korea Tourism Organization	87
39	Expert	90	Improvement	86
40	Ecotourism	81	Revitalization	80
41	Job	81	Expert	77
42	Social Enterprise	80	Outlook	77
43	Plan	79	Overcharging	74
44	Residents	75	Eco-friendly	67
45	Daejeon Fair Tourism	68	Festival	67
46	Youth	60	Tradition	67
47	Eco-friendly	53	Tourist Attraction	67
48	Local Economy	50	Protection	63
49	Brand	46	Souvenir	58
50	Cooperative	45	Plan	57

presents the results of the frequency analysis of the top 50 keywords related to fair tourism for each analytical period. Frequency analysis was conducted to identify the most salient terms in online discourse on fair tourism and to compare the main topics that appeared in each period. During the pandemic continuation and early tourism recovery period, “Region” (404) showed the highest frequency, followed by “Ministry of Culture, Sports and Tourism” (373), “China” (325), “Support” (319), “Program” (300), “Culture” (281), “Development” (248), “Project” (242), “Tourist Destination” (229), and “Contest” (221). These keywords indicate that fair tourism discourse during this period was strongly associated with regional tourism, public policy support, government-led tourism projects, and tourism program development. In particular, the high frequency of “Region,” “Support,” “Program,” “Development,” and “Project” indicates that fair tourism was mainly discussed in relation to regional tourism recovery and policy-based tourism initiatives. The appearance of “Ministry of Culture, Sports and Tourism” and “Korea Tourism Organization” also shows that public-sector policy projects and institutional promotion were visible components of fair tourism discourse. In addition, keywords such as “Environment,” “Fair Trade,” “Fairness,” “Fair Ecotourism,” “Ecotourism,” “Social Enterprise,” “Residents,” “Youth,” “Local Economy,” and “Cooperative” indicate that the discourse also included themes related to sustainability, the social economy, resident participation, and local economic recovery.

During the post-endemic tourism restructuring period, “Foreigners” (564) recorded the highest frequency, followed by “China” (343), “Daejeon” (314), “Program” (295), “Ministry of Culture, Sports and Tourism” (273), “Culture” (244), “Tourists” (243), “Korea” (239), “Tourist Destination” (193), and “Busan” (187). Compared with the first period, the most notable change was the increased visibility of keywords related to inbound tourism, international tourism, and overseas tourism markets, such as “Foreigners,” “China,” “Japan,” “United States,” and “Global.” This suggests that, after the endemic transition, fair tourism discourse became more closely associated with the recovery of international mobility and the reopening of tourism markets. In addition, keywords related to tourism consumption and service experiences, including “Hotel,” “Accommodation,” “Reservation,” “Price,” “Service,” “Tour,” “Experience,” and “Market,” appeared more prominently during this period. Keywords such as “Unfairness,” “Controversy,” “Overcharging,” and “Improvement” also indicate that issues related to price fairness, service quality, market order, and tourist complaints became more visible in online discourse.

A comparison of the two periods shows a shift in the main topics associated with fair tourism. In the pandemic continuation and early tourism recovery period, fair tourism was frequently discussed in relation to public policy, regional recovery, ecotourism, the social economy, and community participation. By contrast, in the post-endemic tourism restructuring period, the discourse became more strongly associated with inbound tourism recovery, tourism market reopening, accommodation and service use, tourist experiences, and price and service fairness.

Overall, the frequency analysis shows that the most frequently mentioned keywords changed from policy- and community-oriented terms in the first period to market-, consumption-, and fairness-related terms in the second period. However, frequency analysis only identifies which keywords appeared frequently; it does not explain how those keywords were structurally connected within the semantic network. Therefore, the following sections examine the centrality, network structure, and CONCOR clusters of fair tourism keywords in order to identify the relational and structural characteristics of online discourse on fair tourism.

4.2. Centrality Analysis of Fair Tourism Semantic Networks

Table 3. Degree and Closeness Centrality of Fair Tourism Keywords by Analytical Period.

Rank	Period 1: Pandemic Continuation and Early Tourism Recovery Period			Period 2: Post-Endemic Tourism Restructuring Period
	Keyword	Degree	Closeness	Keyword	Degree	Closeness
1	Region	0.133	1.000	Foreigners	0.133	0.990
2	Ministry of Culture, Sports and Tourism	0.047	0.980	China	0.048	0.959
3	China	0.045	0.949	Daejeon	0.060	0.959
4	Support	0.106	1.000	Program	0.092	0.980
5	Program	0.058	1.000	Ministry of Culture, Sports and Tourism	0.054	0.990
6	Culture	0.165	1.000	Culture	0.170	0.990
7	Development	0.084	0.990	Tourists	0.065	0.980
8	Project	0.143	1.000	Korea	0.115	1.000
9	Tourist Destination	0.035	1.000	Tourist Destination	0.047	0.990
10	Contest	0.026	0.939	Busan	0.020	0.857
11	Seoul	0.048	0.990	Japan	0.040	0.980
12	Korea	0.111	1.000	History	0.115	0.990
13	Operation	0.058	1.000	Hotel	0.053	1.000
14	World	0.032	0.980	United States	0.030	0.980
15	Gyeonggi-do	0.019	0.959	Development	0.012	0.918
16	Information	0.044	0.990	News	0.034	0.980
17	COVID-19	0.034	1.000	Industry	0.015	0.898
18	Tourists	0.041	0.990	Seoul	0.058	0.990
19	Revitalization	0.044	0.990	Project	0.027	0.969
20	Japan	0.014	0.898	Market	0.023	0.918
21	Economy	0.071	0.990	Experience	0.069	0.980
22	Jeju	0.043	0.980	Planner	0.051	0.949
23	Education	0.054	1.000	Event	0.050	0.990
24	Environment	0.063	1.000	Nationwide	0.059	1.000
25	Korea Tourism Organization	0.019	0.980	Education	0.016	0.949
26	Company	0.100	1.000	Contest	0.074	0.990
27	Hotel	0.026	0.969	Global	0.026	0.929
28	Industry	0.082	0.990	Service	0.077	0.980
29	Event	0.006	0.908	Unfairness	0.016	0.949
30	Content	0.031	0.980	Accommodation	0.025	0.969
31	United States	0.013	0.949	Jeju	0.017	0.888
32	History	0.039	1.000	Fairness	0.043	0.990
33	Fair Trade	0.013	0.939	Controversy	0.007	0.959
34	Promotion	0.038	0.990	Reservation	0.027	0.990
35	Fairness	0.014	0.918	Price	0.020	0.969
36	Fair Ecotourism	0.019	0.847	Tour	0.001	0.602
37	Gangwon-do	0.011	0.888	Fair Trade	0.007	0.878
38	Fostering	0.025	0.959	Korea Tourism Organization	0.023	0.980
39	Expert	0.014	0.969	Improvement	0.011	0.969
40	Ecotourism	0.032	0.949	Revitalization	0.009	0.888
41	Job	0.017	0.939	Expert	0.010	0.929
42	Social Enterprise	0.014	0.949	Outlook	0.004	0.806
43	Plan	0.043	0.990	Overcharging	0.033	0.980
44	Residents	0.047	1.000	Eco-friendly	0.038	0.980
45	Daejeon Fair Tourism	0.011	0.755	Festival	0.028	0.786
46	Youth	0.012	0.959	Tradition	0.012	0.939
47	Eco-friendly	0.008	0.939	Tourist Attraction	0.012	0.959
48	Local Economy	0.013	0.918	Protection	0.007	0.898
49	Brand	0.006	0.959	Souvenir	0.009	0.929
50	Cooperative	0.013	0.959	Plan	0.013	0.949

presents the results of degree centrality and closeness centrality analyses of fair tourism keywords for the pandemic continuation and early tourism recovery period and the post-endemic tourism restructuring period. While the frequency analysis identified the most frequently mentioned keywords, centrality analysis was conducted to examine which keywords occupied structurally important positions within the semantic network.

During the pandemic continuation and early tourism recovery period, the degree centrality results showed relatively high values for “Culture” (0.165), “Project” (0.143), “Region” (0.133), “Korea” (0.111), “Support” (0.106), “Company” (0.100), “Development” (0.084), “Industry” (0.082), “Economy” (0.071), and “Environment” (0.063). These results indicate that keywords related to culture, policy projects, regions, public support, industry, the economy, and the environment were directly connected to many other keywords within the fair tourism semantic network. In particular, “Region,” “Support,” “Project,” and “Development” occupied important positions in the network, suggesting that fair tourism discourse during this period was structurally linked to regional tourism recovery and public policy-based tourism initiatives. The high degree centrality of “Culture” and “Environment” also shows that cultural values and environmental sustainability were not peripheral topics, but were directly connected to other major keywords in the network. Meanwhile, keywords such as “Fair Trade,” “Fairness,” “Fair Ecotourism,” “Ecotourism,” “Social Enterprise,” “Residents,” “Local Economy,” and “Cooperative” showed relatively lower degree centrality values, but they remained thematically important because they reflected the value-oriented and community-based dimensions of fair tourism.

In terms of closeness centrality during the pandemic continuation and early tourism recovery period, “Region,” “Support,” “Program,” “Project,” “Tourist Destination,” “Korea,” “Operation,” “COVID-19,” “Education,” “Environment,” “Company,” “History,” and “Residents” recorded values of 1.000. These keywords were highly accessible within the overall network, meaning that they were connected to other keywords through short semantic paths. This result suggests that the first-period network was organized around a broad set of policy-, region-, destination-, environment-, and resident-related keywords. Therefore, the closeness centrality results complement the degree centrality results by showing that regional tourism recovery, public support, environmental sustainability, and residents were not only frequently connected to other keywords, but also positioned close to the overall discourse structure.

During the post-endemic tourism restructuring period, the degree centrality results showed high values for “Culture” (0.170), “Foreigners” (0.133), “Korea” (0.115), “History” (0.115), “Program” (0.092), “Service” (0.077), “Contest” (0.074), “Experience” (0.069), “Tourists” (0.065), “Daejeon” (0.060), “Nationwide” (0.059), “Seoul” (0.058), “Ministry of Culture, Sports and Tourism” (0.054), “Hotel” (0.053), “Planner” (0.051), and “Event” (0.050). Compared with the first period, the second-period network showed stronger direct connections around inbound tourism, cultural and historical resources, tourist experiences, services, accommodation, regional programs, and tourism planning. The high degree centrality of “Foreigners” indicates that foreign tourists and international tourism recovery became structurally important within online fair tourism discourse after the endemic transition. In addition, the centrality of “Service,” “Experience,” “Hotel,” and “Tourists” suggests that tourism consumption and service-related issues became more directly connected to the fair tourism semantic network. In terms of closeness centrality during the post-endemic tourism restructuring period, “Korea,” “Hotel,” and “Nationwide” recorded values of 1.000, while “Foreigners,” “Ministry of Culture, Sports and Tourism,” “Culture,” “Tourist Destination,” “History,” “Seoul,” “Contest,” “Fairness,” and “Reservation” showed high values of 0.990. This indicates that these keywords were closely connected to the overall semantic network of fair tourism after the endemic transition. These results indicate that national-level tourism, accommodation, inbound tourism, public institutions, cultural tourism, destination-related terms, fairness, and reservation-related issues were positioned close to the overall semantic network. In particular, the high closeness centrality of “Hotel,” “Reservation,” and “Fairness” suggests that tourism service use and fairness-related issues became more accessible within the overall discourse structure after the reopening of the tourism market.

A comparison of the two periods shows that the structural center of online fair tourism discourse changed. In the pandemic continuation and early tourism recovery period, structurally important keywords were mainly related to regional tourism recovery, public policy projects, environmental sustainability, and residents. In the post-endemic tourism restructuring period, keywords related to foreign tourists, tourist experiences, services, hotels, reservations, fairness, and market issues became more central. These findings indicate that the differences observed in the frequency analysis were also reflected in the structural positions of keywords within the semantic network.

The semantic network visualizations in Figure 2 and Figure 3 provide a visual supplement to the centrality results. Figure 2 shows that the first-period network was organized around “Region,” “Support,” “Project,” “Development,” “Environment,” and “Residents,” indicating a policy- and community-oriented discourse structure. Figure 3 shows that the second-period network was more strongly associated with “Foreigners,” “Tourists,” “Hotel,” “Service,” “Experience,” “Reservation,” “Price,” “Overcharging,” and “Unfairness,” indicating a discourse structure more closely related to inbound tourism recovery, tourism service use, price fairness, and market order. Overall, the centrality and visualization results suggest that online fair tourism discourse shifted from a recovery- and policy-oriented semantic structure to a more market-, service-, and fairness-oriented semantic structure after the endemic transition. These findings therefore indicate a structural shift in online fair tourism discourse during the post-pandemic tourism transition.

4.3. CONCOR Analysis of Fair Tourism Semantic Clusters

CONCOR analysis was conducted to identify the sub-discourse clusters within the semantic networks of fair tourism. Based on structural equivalence, keywords with similar relational patterns were grouped into clusters. The semantic networks for both the pandemic continuation and early tourism recovery period and the post-endemic tourism restructuring period were divided into four clusters. The cluster names were determined according to the representative keywords, semantic commonality, and relational characteristics of each group, following the interpretation procedure described in Section 3.4. The results are presented in

Table 4. CONCOR Analysis of Fair Tourism Semantic Clusters by Analytical Period.

Period	Cluster	Cluster Title	Representative Keywords
Period 1: Pandemic continuation and early tourism recovery period	Cluster 1	Public Policy and Regional Tourism Support	Region, Ministry of Culture, Sports and Tourism, Korea Tourism Organization, Support, Project, Program, Promotion, Plan
	Cluster 2	Fair Ecotourism and Sustainable Local Programs	Fair Ecotourism, Ecotourism, Eco-friendly, Daejeon Fair Tourism, Contest, Program
	Cluster 3	Social Economy and Community-Based Recovery	Social Enterprise, Cooperative, Residents, Youth, Job, Local Economy, Revitalization, Economy, Fostering
	Cluster 4	Culture, Tourism Industry, and International Context	Culture, History, Tourists, China, Japan, United States, World, Hotel, Industry, Company, Fair Trade, Fairness
Period 2: Post-endemic tourism restructuring period	Cluster 1	Inbound Tourism and Global Tourism Market Recovery	Foreigners, China, Japan, United States, Global, Korea, Tourists, Nationwide
	Cluster 2	Tourism Consumption and Service Fairness	Hotel, Accommodation, Reservation, Service, Price, Overcharging, Unfairness, Controversy, Improvement
	Cluster 3	Cultural Experience and Destination Attractions	Culture, History, Tradition, Festival, Souvenir, Tourist Attraction, Tourist Destination, Experience, Tour
	Cluster 4	Destination Management and Sustainable Tourism Governance	Ministry of Culture, Sports and Tourism, Korea Tourism Organization, Project, Plan, Planner, Expert, Protection, Eco-friendly, Revitalization, Market, Fairness

. For the pandemic continuation and early tourism recovery period, the four clusters were interpreted as public policy and regional tourism support, fair ecotourism and sustainable local programs, social economy and community-based recovery, and culture, tourism industry, and international context. Cluster 1, Public Policy and Regional Tourism Support, included keywords such as “Region,” “Ministry of Culture, Sports and Tourism,” “Korea Tourism Organization,” “Support,” “Project,” and “Program.” The concentration of public institutions, regional terms, and project-related keywords indicates that this cluster represented the institutional and policy-oriented dimension of fair tourism discourse. It reflects online discussions concerning regional tourism initiatives, public support programs, and government-led tourism projects during the first period. Cluster 2, Fair Ecotourism and Sustainable Local Programs, centered on keywords such as “Fair Ecotourism,” “Ecotourism,” “Eco-friendly,” “Daejeon Fair Tourism,” and “Contest.” This cluster connected fair tourism with environmentally oriented tourism practices, regional fair tourism programs, and contest-based initiatives. The keyword composition indicates that sustainability and regionally implemented ecotourism programs constituted a distinct sub-discourse within the first-period network. Cluster 3, Social Economy and Community-Based Recovery, consisted of keywords such as “Social Enterprise,” “Cooperative,” “Residents,” “Youth,” “Job,” “Local Economy,” and “Revitalization.” This cluster linked fair tourism with the social economy, resident participation, youth employment, local economic issues, and regional revitalization. Its relational structure suggests that community participation and local economic recovery were important components of online fair tourism discourse during the pandemic continuation and early tourism recovery period. Cluster 4, Culture, Tourism Industry, and International Context, included keywords such as “Culture,” “History,” “Tourists,” “China,” “Japan,” “United States,” “World,” “Hotel,” “Industry,” “Company,” “Fair Trade,” and “Fairness.” This cluster combined cultural and historical tourism resources with tourism industry actors, international tourism references, and fairness-related concepts. The coexistence of “Fair Trade” and “Fairness” with industry- and market-related keywords indicates that fairness-related meanings were articulated alongside broader cultural, industrial, and international tourism issues.

Overall, the first-period CONCOR results show that online fair tourism discourse consisted of several differentiated but interconnected sub-discourses. Public policy and regional support formed one major dimension, while fair ecotourism, social economy and community recovery, and cultural and international tourism contexts constituted additional thematic structures. These findings describe the relational organization of online discourse during the first period and do not directly demonstrate changes in actual tourism practices or policy outcomes.

For the post-endemic tourism restructuring period, the four clusters were interpreted as inbound tourism and global tourism market recovery, tourism consumption and service fairness, cultural experience and destination attractions, and destination management and sustainable tourism governance. Cluster 1, Inbound Tourism and Global Tourism Market Recovery, included keywords such as “Foreigners,” “China,” “Japan,” “United States,” “Global,” “Korea,” “Tourists,” and “Nationwide.” This cluster indicates that fair tourism discourse after the endemic transition became more closely connected with the recovery of international mobility, inbound tourism, and global tourism markets. Cluster 2, Tourism Consumption and Service Fairness, consisted of keywords such as “Hotel,” “Accommodation,” “Reservation,” “Service,” “Price,” “Overcharging,” “Unfairness,” “Controversy,” and “Improvement.” This cluster shows that fair tourism discourse increasingly incorporated issues of price fairness, service quality, transparent transactions, and tourist complaints in the reopened tourism market. Cluster 3, Cultural Experience and Destination Attractions, included keywords such as “Culture,” “History,” “Tradition,” “Festival,” “Souvenir,” “Tourist Attraction,” “Tourist Destination,” “Experience,” and “Tour.” This cluster suggests that fair tourism was also articulated through cultural experience, local attractions, and destination-based consumption. Cluster 4, Destination Management and Sustainable Tourism Governance, included keywords such as “Ministry of Culture, Sports and Tourism,” “Korea Tourism Organization,” “Project,” “Plan,” “Planner,” “Expert,” “Protection,” “Eco-friendly,” “Revitalization,” “Market,” and “Fairness.” This cluster indicates that public institutions, planning actors, environmental protection, market management, and fairness-related issues formed a governance-oriented sub-discourse within the second-period network.

Overall, the second-period CONCOR results show that online fair tourism discourse after the endemic transition was no longer organized mainly around regional recovery and public support, but was differentiated into sub-discourses related to inbound tourism recovery, tourism consumption and service fairness, cultural destination experiences, and sustainable destination governance. The CONCOR clusters for Period 1 are visualized in Figure 4, while those for Period 2 are shown in Figure 5.

5. Discussion and Conclusions

This original empirical study examined the semantic structure and temporal transformation of online discourse on fair tourism in Korea. The analysis compared two three-year periods: the pandemic continuation and early tourism recovery period, from 1 June 2020 to 31 May 2023, and the post-endemic tourism restructuring period, from 1 June 2023 to 31 May 2026. Online unstructured text data were analyzed using frequency analysis, degree and closeness centrality analysis, semantic network visualization, and CONCOR analysis.

The results show that online fair tourism discourse during the pandemic continuation and early tourism recovery period was primarily associated with regional tourism, public policy support, government-led projects, fair ecotourism, the social economy, resident participation, and local economic recovery. Keywords related to regions, public institutions, support programs, development projects, the environment, social enterprises, residents, and cooperatives were prominent in both their frequency and relational positions. The CONCOR results further differentiated this discourse into clusters related to public policy and regional tourism support, fair ecotourism and sustainable local programs, social economy and community-based recovery, and cultural, industrial, and international tourism contexts.

During the post-endemic tourism restructuring period, online fair tourism discourse became more strongly associated with inbound tourism recovery, international tourism markets, tourism consumption, accommodation and service use, price fairness, and destination management. Keywords related to foreign tourists, hotels, reservations, services, experiences, fairness, and market issues became more visible or structurally central. The corresponding CONCOR clusters reflected inbound tourism and global market recovery, tourism consumption and service fairness, cultural experience and destination appeal, and destination management and sustainable tourism governance.

Taken together, the findings indicate that the shift from policy- and community-oriented discourse to market-, service-, and fairness-oriented discourse reflects changing conditions of tourism recovery. During the pandemic continuation and early tourism recovery period, restricted mobility, public health concerns, and weakened local tourism economies made fair tourism more closely associated with regional recovery, public support, fair ecotourism, the social economy, resident participation, and community resilience. After the endemic transition, however, the reopening of tourism markets and the recovery of inbound tourism increased public attention to tourist experiences, accommodation, reservations, prices, service quality, overcharging, and unfair market practices. This suggests that fair tourism discourse is sensitive to changing tourism governance challenges: fairness was framed mainly around recovery support and local coexistence in a crisis context, but became more closely associated with market transparency, service encounters, price fairness, and destination governance in a recovery context. This shift should be understood as a change in the structure and framing of digitally mediated public discourse, rather than as direct evidence that actual tourist behavior, tourism policies, or institutional practices changed in the same manner.

The findings suggest that fair tourism discourse in the Korean online context extends beyond individual ethical travel or responsible consumption and incorporates broader concerns related to benefit distribution, community participation, market fairness, service quality, destination governance, and tourism justice. By examining relationships among keywords and changes in semantic clusters, this study extends fair tourism research beyond individual-level surveys of attitudes and behavioral intentions. It also demonstrates the value of social big data and semantic network analysis for exploring how tourism concepts are constructed and rearticulated through online public discourse during periods of social and market transition.

5.1. Theoretical Implications

This study offers several theoretical implications. First, it extends the unit of analysis in fair tourism research from individual tourists’ attitudes and behavioral intentions to the relational structure of digitally mediated public discourse. Previous studies have primarily examined perceptions of fair tourism, behavioral intentions, willingness to pay, and destination attitudes through surveys, interviews, and case studies. These approaches are valuable for explaining individual-level psychological and behavioral responses, but they provide limited insight into how fair tourism is socially constructed, circulated, and connected to other tourism issues in online spaces. By analyzing keyword relationships and semantic clusters derived from naturally occurring online text, this study conceptualizes fair tourism not only as an individual attitude or behavioral orientation, but also as a socially constructed discourse formed through relationships among policy, market, community, environmental, and justice-related meanings.

Second, this study provides empirical evidence of the temporal reconfiguration of online fair tourism discourse within the post-pandemic tourism transition. During the pandemic continuation and early tourism recovery period, the discourse was more strongly organized around regional recovery, public support, fair ecotourism, the social economy, and community participation. During the post-endemic tourism restructuring period, it became more closely associated with inbound tourism recovery, accommodation and service use, market fairness, overcharging and unfairness issues, and destination management. These findings do not demonstrate that the underlying concept or actual practices of fair tourism changed in the same manner. Rather, they show that the meanings and issues surrounding fair tourism were rearticulated in online discourse as social conditions and tourism market priorities changed.

Third, the findings clarify the conceptual position of fair tourism in relation to sustainable tourism, responsible tourism, tourism justice, tourism ESG, and destination governance. Fair tourism is not treated in this study as an umbrella concept that replaces these related perspectives. Instead, it is positioned as a context-specific discourse category that foregrounds fairness in the distribution of tourism benefits and costs, stakeholder participation, market practices, policy accountability, and destination management. The first-period emphasis on community recovery and the social economy reflects distributive and participatory concerns, whereas the second-period emphasis on prices, services, market order, and destination governance highlights fairness within tourism consumption and management. In this way, the study links fair tourism to broader debates on distributive, procedural, and recognitional justice while maintaining its practical and policy-oriented character in the Korean tourism context.

Fourth, this study contributes methodologically by combining frequency analysis, degree and closeness centrality analysis, semantic network visualization, and CONCOR analysis within a comparative temporal framework. Each method provides a distinct level of interpretation: frequency analysis identifies salient terms, centrality analysis reveals structurally important and accessible keywords, network visualization shows relational patterns, and CONCOR analysis identifies sub-discourse clusters based on structural equivalence. Applying the same analytical procedures to two equivalent three-year periods enables the study to examine changes in both the content and relational organization of online discourse. Although the empirical data were collected from Korean digital platforms, this approach has broader relevance for tourism discourse research because it can trace how value-laden tourism concepts are rearticulated across changing social, policy, and market contexts. It can also be applied to other national or cultural settings to examine how concepts such as fair tourism, responsible tourism, sustainable tourism, overtourism, or tourism justice become connected to public concerns, policy agendas, and market issues.

5.2. Practical and Policy Implications

This study also offers several practical and policy implications. Because the findings represent patterns of online discourse rather than direct measures of tourist behavior, policy performance, or market conditions, they should be used as diagnostic signals that inform decision-making in combination with visitor surveys, resident feedback, business data, and other empirical evidence.

First, policymakers and local governments can differentiate fair tourism policy priorities according to the stage of tourism recovery. During the pandemic continuation and early tourism recovery period, online fair tourism discourse was strongly associated with regional tourism recovery, public support, the social economy, resident participation, and ecotourism. These findings highlight the relevance of region-based fair tourism programs, partnerships with social enterprises and cooperatives, youth and resident-participatory tourism projects, and environmentally responsible tourism content. During the post-endemic tourism restructuring period, however, price fairness, service quality, overcharging, and fairness in tourist experiences became more visible. Accordingly, future fair tourism policies should not remain limited to financial support and regional revitalization projects, but should also incorporate tourism market fairness, transparent transactions, consumer protection, and service-quality management.

Second, destination management organizations and local tourism organizations should institutionalize fair tourism as a destination governance principle rather than use it only as a promotional message associated with “good travel.” The connections among regions, residents, the local economy, social enterprises, cooperatives, environmental protection, and cultural resources suggest that fair tourism can support the joint management of community benefits, resident participation, environmental conservation, and tourist–resident relationships. From a stakeholder communication perspective, local governments and destination management organizations should establish regular consultation mechanisms involving residents, local businesses, social economy organizations, tourism operators, and tourists when designing tourism products and destination policies. In addition, digital monitoring of online tourism discourse can help identify emerging conflicts or concerns related to overcharging, service dissatisfaction, resident inconvenience, environmental pressure, and unfair tourism practices. However, such monitoring should be integrated into participatory tourism governance rather than used as a purely technical management tool. Issues detected through online discourse should be verified through direct consultation, complaint records, visitor surveys, resident feedback, and business-sector communication, and then translated into agenda-setting, stakeholder dialogue, and corrective policy action.

Third, the management of price and service fairness is crucial in the process of post-endemic tourism market recovery. The online discourse analysis showed that “Price,” “Overcharging,” “Unfairness,” “Controversy,” “Service,” and “Improvement” emerged as major keywords during the post-endemic tourism restructuring period. This suggests that, alongside the recovery of tourism demand, fairness-related issues became more visible in online discourse concerning accommodation, experiences, tours, reservations, and services. Therefore, tourism policy agencies, local governments, and destination management organizations should monitor issues such as overcharging, unfair pricing, discriminatory treatment of tourists, and service complaints. They should also work with local tourism businesses to provide transparent price information, strengthen service quality management, and improve systems for responding to tourist complaints. These measures can help prevent fairness-related issues from escalating into broader destination management problems during the recovery of the tourism market.

Fourth, in the context of inbound tourism recovery, fair tourism should be linked to improvements in inbound tourism readiness. The prominence of keywords such as “Foreigners,” “China,” “Japan,” “United States,” “Global,” and “Tourists” after the endemic transition suggests that online fair tourism discourse became more closely associated with the return of foreign tourists, the recovery of international tourism, overseas tourism markets, and cross-national tourism issues. Accordingly, fair tourism policies should include multilingual information services for foreign tourists, fair and transparent pricing systems, culturally sensitive services, safe tourism environments, and programs that promote mutual respect between local residents and foreign tourists. Local governments, destination management organizations, and tourism businesses should cooperate in implementing these measures and establish accessible multilingual channels through which foreign tourists can obtain information and report service-related problems.

5.3. Limitations and Future Research Directions

Although this study contributes to the literature by examining temporal changes in online discourse on fair tourism, it has three main limitations. First, the data were collected from selected Korean online platforms using the principal Korean search query “공정관광,” translated as “fair tourism” in this study. Although the related expression “공정여행,” or “fair travel,” was reviewed during data cleaning when it appeared in relevant contexts, relying primarily on one search query may have excluded discussions expressed through alternative terminology. Moreover, platform-specific user characteristics, search and exposure algorithms, changes in platform composition, and the relative proportion of public-sector or promotional content may have influenced the observed discourse structure. Therefore, the findings cannot be generalized to all tourists, residents, tourism businesses, or national contexts. Future research should use multiple synonymous search terms, platform-stratified datasets, multilingual corpora, and cross-national comparisons to examine whether similar discourse structures emerge across different digital and cultural contexts.

Second, keyword-based semantic network analysis and document-level co-occurrence analysis have limitations in capturing sentence-level meanings, emotional valence, speaker positions, and the specific contexts in which keywords are used. For example, terms such as “Price,” “Overcharging,” “Unfairness,” and “Controversy” may appear in tourist complaints, media reports, policy discussions, promotional content, or demands for institutional improvement. Their co-occurrence does not by itself establish whether the discourse was positive or negative or whether it reflected actual changes in tourist behavior, market conditions, or institutional governance. Future studies should combine semantic network analysis with sentiment and emotion analysis, topic modeling, stance analysis, contextual text analysis, and manual coding. Surveys and interviews with tourists, residents, tourism businesses, destination management organizations, and policy officials would also help determine how patterns identified in online discourse are related to actual perceptions, practices, and policy demands.

Third, this study divided the six-year dataset into two aggregated three-year periods and constructed the final networks using the top 50 keywords for each period. This comparative design enhanced network readability and enabled consistent comparison, but it may have obscured monthly or yearly fluctuations, event-specific shifts, regional differences, and platform-specific discourse patterns. The network results may also be sensitive to analytical decisions such as the number of keywords included. Future research should conduct monthly or yearly time-series analysis, event-based comparison, regional and platform-specific network analysis, and robustness checks using alternative keyword thresholds. These approaches would provide a more detailed understanding of how fair tourism discourse changes in response to particular tourism events, policy developments, market conditions, and destination-specific issues.