Identifying Local Characteristics for Customized Policy Application Within Rural Areas Using LDA Topic Modelling

Abstract

Rural plans incorporating regional identity are vital for fostering regional revitalization and offering viable policy alternatives. The need for a systematic approach that recognizes both the diversity and shared characteristics of rural areas has become increasingly clear. Although numerous studies have explored rural classification, research examining specific regional characteristics remains limited. Hence, this study aimed to establish a comprehensive standard for developing effective rural plans. To this end, a study was conducted to classify the characteristics of rural areas using topic modeling, which is a text-mining algorithm. An analysis of publications on rural revitalization projects in Korea over the past decade revealed five common factors of success themes across each region. The five success factors, “local cultural experience”, “environment and landscape utilization”, “community activation”, “regional infrastructure development”, and “local economic activation”, should be considered in rural areas with different characteristics when establishing rural plans and policies. This classification of success factors serves as the foundation for establishing rural plans and policies. By applying different weights to the five success factors according to the unique characteristics and conditions of each region, it would open a great number of possibilities to establish more precise and effective customized plans. Future research is required to provide more empirical and broadly applicable results based on the classification framework proposed in this study.

1. Introduction

According to the World Urbanization Prospects published by the UN in 2018, approximately 68% of the global population is expected to live in urban areas by 2050, with about 2.5 billion people predicted to migrate to cities from 2018 to 2050 [1]. This indicates that the urban–rural gap will continue widening due to urbanization, resulting in a steady decline in the rural population [2]. Specifically, as the urban–rural gap grows, access to public infrastructure and services in rural areas diminishes [3], and rapidly urbanizing societies may lead to greater poverty and imbalance [4].

Hence, overcoming the decline in rural areas has become a global challenge [5]. Rural plans are becoming increasingly important as a solution to rural area depopulation. Establishing effective rural plans can enhance the living conditions for residents and create a favorable business environment for companies [6]. One new approach to rural planning considers regional identities [7] and aims at preserving the unique physical, historical, cultural, and ecological characteristics of each rural area to overcome the negative factors affecting rural areas [8,9].

Despite the emergence of various rural plans and policies, the inclusion of regional identity in plans is lacking [10]. In practice, current rural planning tends to simplify complex rural issues into persuasive slogans or quantitative metrics. However, rural areas around the world are highly diverse, making it difficult to define them through standardized quantitative criteria, yet current planning practices often neglect this diversity [11]. In addition, while spatial planning tends to establish mid-to-long-term goals and strategies [12], most plans targeting rural areas often focus on short-term outcomes [13]. This reflects a fixed perception of rural areas as stagnant and undeveloped spaces, which in turn leads to standardized policies that fail to account for local specificities. In Austria, for example, policies that failed to integrate ecological assets and local identities faced resistance from local communities [13]. A similar case occurred in Korea, where approximately 23 billion KRW was invested in constructing a fisheries complex in a mountainous village. However, due to unfavorable transportation conditions and the lack of surrounding tourism infrastructure, the project failed to attract visitors. This case exemplifies a failure to adequately consider regional topographic characteristics, highlighting the need for more localized development strategies.

Various plans and policies can be divided into piecemeal and systemic approaches. The piecemeal approach tends to focus on quantitative growth and mostly follows a simple perspective [14]. In contrast, the systemic approach requires a comprehensive understanding of the organization and is similar to the structural-functional approach [15]. It also plays an important role in addressing the increasing complexity of the multiple stakeholders participating in a single plan [16]. Therefore, to successfully establish rural plans, it is necessary to take a systematic approach based on specialized and detailed knowledge of rural areas [17,18]. To implement such a systematic approach, it is essential first to understand the identity and characteristics of each rural area within a comparable analytical framework. In other words, structurally identifying the similarities and differences among rural regions is a prerequisite for developing tailored plans based on local identity.

Although some studies have been conducted to classify rural areas [19,20], research on systematic approaches that consider the heterogeneity and similarity of each rural area has been limited. If the characteristics of each region are not considered, the development of effective guidelines for rural planning becomes challenging. Therefore, evaluating and classifying each success factor across existing rural revitalization projects is essential for deriving a comprehensive standard.

This study aimed to categorize the key success factors of rural planning by analyzing the regional characteristics of rural areas. To achieve this research objective, unsupervised machine learning was used, and a text-mining methodology based on statistical modeling was adopted. Specifically, the goal was to extract key topics that will contribute to establishing future rural plans from a vast number of rural revitalization cases. Accordingly, this study assumes that there are recurring success factors in rural revitalization and presents the following research questions: (1) What common success factors exist in rural revitalization? (2) How can the extracted topics be applied to future rural planning and policy design?

Well-designed plans can lead to policy improvements in other countries or regions. Therefore, it is necessary to establish optimal standards that can be adapted to diverse contexts. The five key themes identified in this study are expected to serve as practical references for future rural policies and planning, potentially functioning as a foundational framework. In doing so, the study aims to contribute to supporting the sustainable development of rural areas.

2. Literature Review

In recent decades, new policies and plans for rural revitalization have emerged as solutions to rampant urbanization, with regional identity regarded as a strategically valuable asset [21]. Representatively, the Cork 2.0 European Conference on Rural Development, held in 2016 to address the challenges rural Europe faced, stated that it would invest in rural locality and identity and develop approaches that reflect the needs of each region [20]. The conference also indicated that it would systematically review policies that consider rural jobs, development, social welfare, and communities [20].

In this regard, France has implemented a policy that offers differentiated support based on the population size of rural areas [22], whereas the UK has provided subsidies aligned with economic, environmental, and social priorities [23]. Thus, many countries actively promote rural planning. South Korea has implemented various plans for rural development and has evolved in phases in response to economic and social changes.

2.1. The History of Rural Planning in South Korea

The beginning of the rural revitalization policy in the Republic of Korea was the community development project implemented in 1958. After the Korean War, the national economic and social foundations collapsed, and the government promoted rural development by training leaders with the goal of improving agriculture and living environments at the village level [24]. From the 1970s to the late 1990s, a project called the ‘Rural New Village Movement’ was introduced in earnest, and the movement for rural revitalization was systematically expanded.

The Rural New Village Movement was divided into the rural production base sector, rural income increases sector, welfare environment sector, and mental development sector according to the purpose of promotion, and initially, like the community development project, it focused on establishing an agricultural production base [25]. After that, projects for increasing income gradually became the focus, and lower-level projects such as the joint fishing ground creation project, agricultural project distribution structure improvement project, and income increase comprehensive development project were implemented.

However, as industrialization and urbanization progressed, the rural population decreased rapidly. Until the 1970s, the rural population accounted for 57.4% of the total population, but in 2015, it decreased to 18.4% [26]. Accordingly, rural planning also changed. In the past, the focus was on improving productivity and infrastructure in rural areas, but recently, policies to solve the problem of population decline and improve the quality of life of residents have been promoted.

A representative example is the rural development implemented from 2004 to 2017. This industry broke away from the existing top-down development method and emphasized resident participation and implemented village-level projects as a core principle. It was characterized by encouraging residents to directly set development directions reflecting the characteristics of local resources and seeking ways to contribute to increasing income [27].

These policies did not always produce successful results. Although they emphasized resident participation, they were criticized for being planned mainly by the government and external experts [28,29], and there were cases where uniform development methods that did not reflect the characteristics of the region failed to produce economic effects. Therefore, rural policies need to be designed around customized strategies that reflect the unique characteristics of each region [30].

2.2. Heterogeneity and Similarity Investigation in Rural Areas

Classification can be used as an analytical tool to better understand regional characteristics and differences [31,32]. Appropriate classification provides a foundation for analyzing past and present trends, predicting future changes, and establishing flexible policies. A representative study on the subject of classification is the urban–rural classification based on population ratio, which is the basis of the French rural policy mentioned previously [19,31,33]. This distinction has been proven to explain regional differences better than general statistics grouped by administrative districts such as local governments; however, it has limitations in capturing each regional characteristic [34,35]. It has also been criticized for oversimplifying complex regional differences based on only one criterion: population size [34,36]. Rural plans must consider the new and more complicated relationships as society changes [37].

To overcome this limitation, Yi and Son (2022) conducted a study that categorizes rural areas according to four characteristics: preservation, celebration, community, and creativity [38]. Rural planning should be differentiated based on regional characteristics, and previous studies highlight the need to propose concrete directions to achieve sustainability, coexistence, prosperity, and well-being in rural areas [39]. This is significant because it confirmed that systematic spatial plans and research in rural areas are insufficient compared to cities; however, the results were not specific owing to a lack of cases [38].

Meanwhile, a study was conducted to classify rural development types in Yanchi County, located in northern China [40]. This study classified the rural development structure into eight types, including the rural industry development model, rural governance model, traditional village projection, and tourism model, targeting villages in Yanchi County. Although the need for differentiated development strategies considering the natural environment and industrial structure of the region was emphasized through analysis by type, it is limited to a specific county unit rather than a national classification, so there are limitations.

Most of the rural classifications distinguish between urban and rural areas in a broad sense, and even when based on specific characteristics, they often lack sufficient case studies for validation. In particular, many studies rely on narrow criteria such as population or industry type, which can oversimplify complex rural societies [39,41]. For the activation to be successful, it is necessary to go beyond classifying rural areas by population and area and establish criteria based on regional identity [42,43]. Research needs to be conducted to better capture a higher level of diversity and heterogeneity in rural areas [44,45]. To overcome the limitations of these existing studies, this study proposes a multidimensional analysis of rural space and a comprehensive strategy for rural revitalization. It digests a large amount of data and extracts specific regional characteristics using text mining techniques. In addition, while many previous studies remained at the level of conceptual discussion, this study takes a more practical approach. Based on data from specialized village projects, it classifies rural types and identifies key activation factors. Following this classification, the study provides policy implications for each type, offering a potential foundation for future rural policy development.

2.3. Heterogeneity and Similarity of Rural Areas Through Statistical Modeling

Analyzing rural revitalization projects implemented in South Korea over the past decade is a complex task that requires considerable time and often results in low accuracy [46]. To compensate for these problems, text-mining techniques that excel in processing large datasets can be utilized [47]. They are widely used in policy research and to classify substantial amounts of text during the literature review process [48,49]. In the rural field, Chang et al. effectively utilized text mining to examine rural tourism and sustainable development documents, classifying them into economic, environmental, and social based on shared themes [50]. Tang et al. used text-mining technology to analyze papers on sustainable development and tourism [51]. In this way, text-mining techniques not only improve the accuracy of research results but also simplify the review process, paving the way for more informed decision-making. In this study, topic modeling was used for text mining. First, the characteristics of each rural area were classified through k-means clustering analysis, and the success factors of each type were derived through topic modeling.

Topic modeling is a useful technique for extracting key information related to potential topics from large amounts of text data [52,53]. For example, text mining was used together with topic modeling in a paper that studied the nature of rural development in the Rural New Village Movement. After identifying the overall content of the Rural New Village Movement implemented in 1980 through text mining, the essence was identified through topic modeling analysis [53].

The methodology that combines text mining and topic modeling is a technique widely used in social science today, and this study also uses topic modeling for text mining. First, the characteristics of each rural area are classified through k-means clustering analysis, and then the success factors of each type are derived through topic modeling. Topic modeling helps establish a stronger foundation by automatically grouping terms assumed to be semantically related and is also beneficial for sifting through large amounts of data [54]. Specifically, it is widely used in research to identify which topics attract more customers and what trends exist in texts containing documents, comments, reviews, and so on [54].

The reasons for selecting text mining and topic modeling as a research methodology are summarized as follows: (1) It is highly useful because it can discover core topics from the text. (2) It can digest large amounts of textual data. (3) The beta value can be used to determine the relative contribution of words to publications. Topic modeling checks the effectiveness of rural revitalization projects and contributes to the discovery of elements necessary for framing effective rural planning guidelines.

Therefore, this study aims to analyze various text data on rural revitalization projects using text mining techniques and then extract success factors for each type of project, thereby contributing to the provision of practical guidelines for rural revitalization.

3. Methodology

As rural areas encompass various sectors, rural planning must shift from a general focus on agriculture to a strategy that acknowledges diverse economic potential and needs [55]. This study aimed to analyze the socioeconomic performance achieved through rural revitalization projects, classify them according to the characteristics of each region, and utilize these as key factors to consider in future rural planning. It is necessary to enhance the objectivity of the analysis results and the diversity of interpretations to analyze the relative importance of regional characteristics. Thus, in this study, unsupervised learning was used for topic modeling based on statistical modeling.

The study was conducted in the following steps: (1) Approximately ten years’ worth of publications (e.g., policy statements and government reports) issued in Korea were collected and reviewed in depth. (2) Keywords were extracted through text-mining technology following the data preprocessing process. (3) Cluster analysis was used to categorize them, and topic modeling by type was used to derive the performance elements that led to rural revitalization projects.

3.1. Data Collection and Preprocessing

In the UK, the Department for Environment, Food and Rural Affairs, and in France, the Ministère de l’Agriculture et de l’Alimentation, formulate and implement various policies related to rural development. Similarly, in South Korea, the Ministry of Agriculture, Food and Rural Affairs (MAFRA) is the highest agency responsible for rural planning and revitalization, income stability, and welfare promotion. MAFRA oversees various rural revitalization projects, including a project to select best practices for rural areas across the country. Therefore, this study analyzed the success factors of villages recognized as excellent in rural revitalization by government agencies, including MAFRA. About best practices, this study used data obtained from the media (e.g., news, newspapers), official government data (e.g., policy reports), and data exchanges (e.g., public data platforms) as sources and collected 479 data cases over 10 years from 2014 to 2023. Owing to the large amount of available data, the ability to analyze it well was crucial. Python 3.12.2 and Visual Studio Code 1.89.1 were used to preprocess a large amount of text data into a form that could be explored. Python’s regular built-in expressions were employed in this analysis to eliminate English characters, numbers, symbols, special characters, and stop words, excluding Korean. Subsequently, a term-document matrix was created using TF-IDF to transform the data into an analyzable format. The term frequency-inverse document frequency (TF-IDF) algorithm is a traditional feature extraction technique that is simple to implement yet powerful, and it effectively identifies important terms within a text corpus [56,57]. In addition, tokenization was conducted based on nouns using an okt morphological analyzer from the Konly library.

3.2. Data Analysis Using K-Means Clustering and LDA Topic Modeling

Based on the data in an analyzable form, K-means cluster analysis and Latent Dirichlet Allocation (LDA) topic modeling were performed to derive the key factors of each cluster. First, K-means cluster analysis is one of the most used unsupervised learning algorithms, which classifies data with similar characteristics into k predefined clusters [58,59,60]. In this study, the elbow method, a commonly used technique for determining the optimal number of clusters, was applied. Based on the elbow curve, the optimal value of k was set to 5, where the inertia begins to level off and the slope of the curve becomes noticeably flatter. The validity of the chosen number of clusters was then evaluated using metrics such as the silhouette coefficient. Next, dimensionality reduction followed by Principal Component Analysis (PCA) was performed to evaluate the clustering results and identify the centroid of each cluster. Additional visualization techniques were also applied for clearer classification. Afterward, keywords for each cluster and the entire document set were extracted using Python’s built-in functions and expressions. To identify the factors influencing the success of the specialized village project, LDA topic modeling was conducted. LDA is a form of topic modeling technique that assumes each document is a mixture of multiple topics, and each topic is represented by a probability distribution over words [61,62]. It enables the automatic extraction of latent topics from large collections of text [63]. For this, a custom text preprocessing function was developed, a word-document matrix was generated, and the LDA model was trained. The number of topics was determined by calculating and visualizing the perplexity score for each candidate topic count. Considering the slope of perplexity decrease and the interpretability of the topics, the optimal number of topics was set to 5. Based on this, the final LDA modeling results were obtained.

4. Results

4.1. Five Factors for Rural Success

After conducting a K-means cluster analysis, keywords defining the characteristics of each cluster were derived to classify the rural revitalization projects. LDA topic modeling requires setting an appropriate number of topics in advance. In this study, the optimal number of topics was determined using the elbow method, which identifies the point at which the inertia value stabilizes as the number of clusters increases. Additionally, the number of topics was set to five by comprehensively considering the congestion index and the number of documents.

In determining the number of topics, not only the perplexity score and the elbow method were considered, but also the interpretability of the results played a crucial role. When the number of topics exceeded five, the semantic cohesion among keywords within each topic weakened, and issues such as topic redundancy and excessive fragmentation emerged. Conversely, with fewer than five topics, each topic became overly broad, making practical policy application difficult. Therefore, setting the number of topics to five was deemed appropriate, as it maintained internal coherence within each topic while clearly reflecting the key themes related to rural revitalization.

After classifying the clusters into five, we reviewed the success factors of rural revitalization projects by deriving the main topics of all rural projects and each cluster using topic modeling. The results of the LDA topic modeling with the number of topics set to five for all cases in rural areas are shown in

Table 1. LDA topic modeling results deriving success factors.

Topic 1: Local Cultural Experience		Topic 2: Utilization of Environment and Landscape		Topic 3: Community Activation		Topic 4: Establishment of Local Infrastructure		Topic 5: Local Economic Activation
Word	Beta	Word	Beta	Word	Beta	Word	Beta	Word	Beta
Residents	0.0522	Residents	0.0540	Residents	0.0699	Local	0.0327	Experience	0.0941
Culture	0.0354	Landscape	0.0205	Culture	0.0280	Residents	0.0296	Program	0.0251
Business	0.0191	Environment	0.0190	Business	0.0262	Culture	0.0228	Operation	0.0202
Experience	0.0172	Project	0.0161	Local	0.0188	Business	0.0213	Development	0.0191
Community	0.0150	Making	0.0131	Operation	0.0162	Operation	0.0194	Income	0.0143
Activity	0.0142	Community	0.0129	Activity	0.0156	Activation	0.0186	Festival	0.0135
Income	0.0130	Nature	0.0118	Welfare	0.0135	Program	0.0185	Residents	0.0124
Program	0.0128	Composition	0.0107	Community	0.0127	Education	0.0170	Making	0.0122
Local	0.0120	Activity	0.0107	Program	0.0110	Center	0.0152	Utilization	0.0096
Festival	0.0112	Festival	0.0103	Activation	0.0086	Space	0.0145	Recreation	0.0080
Operation	0.0111	Participation	0.0076	Returning to Farming	0.0086	Hub	0.0140	Production	0.0074
Participation	0.0105	Trash	0.0076	Club	0.0083	Facilities	0.0135	Sales	0.0071
Tradition	0.0094	Development	0.0074	Festival	0.0080	Utilization	0.0127	Agricultural Products	0.0070
Art	0.0091	Change	0.0065	Rural Relocation	0.0077	Support	0.0125	Business	0.0069
Utilization	0.0087	Returning to Farming	0.0063	Participation	0.0074	Mastermind	0.0106	Visitors	0.0064
Development	0.0069	Resource	0.0062	Education	0.0073	Community	0.0102	Ecology	0.0060
Facilities	0.0061	Promotion	0.0061	Facilities	0.0061	Welfare	0.0095	Tourism	0.0058
Harmony	0.0056	Meeting	0.0057	Composition	0.0060	Service	0.0088	Mud flat	0.0057
Exploitation	0.0054	Ecology	0.0055	Development	0.0059	Picture book	0.0087	Local	0.0056
Resources	0.0053	Income	0.0054	Events	0.0058	Society	0.0086	Special Products	0.0048

. The beta values shown in Table 1 reflect the extent to which each word contributes to various topics. Specific topics can be formed from words with relatively high beta values, distinguishing them from other topics. The factors influencing the success of rural revitalization projects in rural areas include “local cultural experience”, “environment and landscape utilization”, “community activation”, “regional infrastructure development”, and “local economic activation”.

The results reveal that words such as residents, projects, participation, and region were emphasized in most of the topics. When the same words appear repeatedly, it means that they have important meanings throughout the rural revitalization project. Excluding general words, we derived specific features for each topic based on the words that could be differentiated.

Topic 1, which relates to local cultural experiences, emphasized words such as culture, experience, programs, and festivals. This indicates that developing programs to engage with local content and traditions has impacted the success of rural revitalization projects. Furthermore, the inclusion of words such as income and development underscores the necessity of managing the festival in a way that positively affects the local economy rather than merely organizing it.

Topic 2 showed a high frequency of words such as landscape, environment, creation, nature, and composition. This means that many projects actively utilize each rural area’s unique natural environment and landscape. Thus, it can be inferred that the interest of outsiders was attracted, and the project was successfully led.

In Topic 3, the word contribution of residents was very high, and words such as culture, welfare, community, and program were prominent, confirming that the expansion and development of local communities was a trend. It also emphasized the importance of communication and community strengthening based on the local culture. What differentiates it from other topics is that it includes returning to farming and rural relocation, which was interpreted as an environmental improvement project to support the return to farming and rural areas, thus having a positive effect on rural revitalization.

Topic 4 is characterized by words such as business, program, education, and center, which can be regarded as an issue related to establishing a foundation for strengthening the capacity of the local community. This means that actively utilizing local facilities and providing education to residents are important.

Topic 5 emphasized experience, development, income, and utilization, confirming that the development of products and services can contribute to revitalizing the local economy. It also highlights that a project can be evaluated as successful only if it contributes to creating jobs or increasing residents’ income.

Through topic modeling of the entire project, the main topics necessary for rural revitalization were identified. Subsequently, a word cloud was implemented to visualize the important words and issues (Figure 1). Typically, when analyzing text, a word cloud is implemented based on simple frequency. However, here, visualization was performed based on the importance of words derived through topic modeling. As a result of the word cloud visualization, the importance of words such as resident, local, program, and culture was found to be very high. As confirmed by the topic modeling results above, this emphasizes that issues such as resident participation, experience, utilization of local culture, and program diversification affect project implementation performance, as these words have a common effect on each topic.

4.2. Policy Direction for Rural Revitalization in Korea

In this study, we will discuss how the five success factors derived through topic modeling, ‘local cultural experience’, ‘environment and landscape utilization’, ‘community activation’, ‘regional infrastructure development’, and ‘local economic activation’, can be applied to the revitalization project of rural villages in Korea. Through this, we will analyze how each success factor can be implemented in actual policies and projects and suggest strategic directions for sustainable rural revitalization. To this end, we conducted an analysis focusing on excellent rural village cases where the five success factors were applied. We selected excellent village cases by referring to the MAFRA project, which is the most influential project in the rural revitalization sector in Korea. This project selects villages that contribute to regional development and the improvement of residents’ quality of life through cooperation between residents and local governments. There are six criteria for selecting specific excellent village cases. These are performance, regional suitability, innovation, resident participation rate, flexible response to external changes, and implementation of regional identity. These criteria reflect the success factors of rural revitalization in various ways, and the selected cases can serve as important basic data for suggesting a sustainable rural development model.

4.2.1. Topic 1: Local Cultural Experience

Topic 1 shows that experiential programs utilizing local cultural characteristics are effective strategies for rural revitalization. A village in Yeoncheon-gun, Gyeonggi-do, was recognized as an excellent example of rural revitalization in 2021. This village has been running an experiential program using cucumbers, a local agricultural product, since 2009. Representative programs include cucumber soap making and cucumber cooking experiences. By utilizing the residents’ farms as a rural experience space, it was able to proceed smoothly, and it is being used as an experiential tourism resource while informing visitors of the value of local agricultural products. In addition, the village holds a cucumber festival every year, and more than 18,000 tourists visit the small rural village with 266 households every year. Through this, farmers are reaping the effects of increasing their actual income and creating jobs. The village above is a representative case that well demonstrates the main keywords derived from Topic 1, such as residents, culture, business, and experience. Through this, we can see that experiential programs based on local culture can promote the attraction of visitors and contribute to the revitalization of the local economy and rural revitalization.

4.2.2. Topic 2: Environment and Landscape Utilization

Topic 2 demonstrates that utilizing local natural environments and landscapes is an effective strategy for rural revitalization. Boeun-gun, Chungcheongbuk-do, which was recognized as an exemplary case of rural revitalization in 2020, successfully developed an eco-friendly village. Situated on a mountainside at an altitude of 877 m, the village boasts a pristine natural landscape and a 200-year-old pine forest. The residents actively preserved this unique natural environment and incorporated it into the village’s identity. To maintain environmental sustainability, the community refrained from operating livestock farms and implemented herbicide-free farming practices. Additionally, residents voluntarily established a “‘Landscape Conservation Residents’ Agreement” and expanded environmental protection facilities. These efforts not only contributed to landscape conservation but also attracted more visitors, with annual tourist numbers steadily increasing since 2016, when approximately 6000 visitors were recorded. This village serves as a representative case that encapsulates key themes derived from Topic 2, such as residents, landscape, nature, and environment. The proactive environmental conservation efforts of the local community have played a crucial role in enhancing the village’s ecological value and have contributed to rural revitalization.

4.2.3. Topic 3: Community Activation

Topic 3 highlights the significance of developing and expanding local communities as a key factor in rural revitalization. Even in rural areas that lack distinctive natural landscapes or cultural assets, community-driven initiatives can contribute to regional development. A notable example is a village in Suncheon-si, Jeollanam-do, which was recognized as an exemplary case of rural revitalization in 2023. The village has actively implemented programs to support urban-to-rural migrants, aiming to strengthen community capacity. These initiatives include economic support measures and job creation programs, providing both practical assistance and a foundation for long-term settlement. As a result, the village population, which stood at only 24 in 2009, had tripled by 2023, while the average income of residents increased more than fivefold. The increased economic resources have been reinvested into cultural and welfare programs, contributing to the enhancement of residents’ quality of life and community cohesion. This case exemplifies key themes derived from Topic 3, including residents, culture, rural migration, and welfare. Furthermore, the strengthened sense of community and social integration among residents has played a pivotal role in fostering a more sustainable and livable rural environment.

4.2.4. Topic 4: Regional Infrastructure Development

Topic 4 highlights the necessity of establishing a foundation for the development of local communities. A notable example is a village in Seogwipo-si, Jeju-do, which was recognized as an excellent case of rural revitalization in 2014. Although the village initially had a limited presence in Jeju-do, communication among residents began to flourish in 2009 following the establishment of cultural and welfare infrastructure. The creation of a cultural school for residents served as a catalyst for expanding community activities. The village’s cultural school currently offers 13 programs, including dance and singing, with approximately 355 residents participating as members. These cultural initiatives contributed to the revitalization of the local community by fostering social engagement and increasing residents’ interest and participation in rural revitalization efforts. Furthermore, residents voluntarily formed organizations such as senior citizens’ associations and village development committees to promote customized education programs, which led to the region’s spontaneous development. This case exemplifies key concepts derived from Topic 4, including business, programs, education, and space, demonstrating that tailored educational initiatives and infrastructure development can drive positive transformations in rural areas.

4.2.5. Topic 5: Local Economic Activation

Topic 5 underscores the significance of developing practical products and services as a fundamental strategy for revitalizing the local economy. A notable example is a village in Hapcheon-gun, Gyeongsangnam-do, which was recognized as an exemplary case of rural revitalization in 2016. The village initiated an agricultural product processing business in 2005, leveraging soybeans—a locally produced specialty—to establish its own brand. A key factor in its success was the direct management of the entire production, processing, and sales process by residents. This approach enabled the supply of high-quality processed products at competitive prices, thereby contributing to economic revitalization. As a result, the village experienced steady income growth and successfully secured approximately 10,000 direct transaction customers. Furthermore, a portion of the village’s revenue was reinvested into the local community, facilitating the operation of a communal meal program that provided free lunches to all residents daily. Through this initiative, the village effectively established a sustainable community model free from economic hardship. This case exemplifies the key concepts associated with Topic 5, including experience, operation, development, income, and utilization, demonstrating the critical role of strengthening the local economic foundation and promoting community-driven operations.

Through the above cases, we have confirmed the efforts of each type of rural area to voluntarily secure competitiveness. Through this, we can confirm the importance of best understanding the characteristics, history, and culture of the region. At this time, referring to the five success factor classifications presented in this study can help to more systematically analyze and reflect regional characteristics and differences. Based on this, a table summarizing the correspondence between the success factors derived through LDA topic modeling and actual best practices is presented below.

Table 2. Policy response plan for each success factor of rural revitalization.

Success Factors	Case Application
Local cultural experience	Experience programs using local specialty cucumbers and preservation of cultural identity
Utilization of environment and landscape	Environment-friendly village development through pine forest preservation and resident-led landscape management
Community activation	Support for returning farmers and rural migrants, improvement of community quality of life through cultural and welfare programs
Establishment of local infrastructure	Operation of cultural schools and establishment of related facilities, implementation of customized education programs for residents
Local economic activation	Development of processed agricultural product brands, operation of direct sales platforms, and profit-sharing systems

visually demonstrates how each success factor contributed to regional activation, rather than serving merely as a keyword.

5. Discussion

This study employed text mining and topic modeling techniques to examine the socioeconomic outcomes of rural development projects and discussed how these findings could be applied to actual rural areas. After reviewing and deriving indicators from rural development programs in Korea over the past ten years, keywords representing the characteristics of each cluster were identified to extract the topics implied by the clusters.

Table 3. Rural revitalization topics and their corresponding success factors and policy applications.

Section	Success Factors and Policies Relation
General criteria	- Operation of local cultural experience programs
	- Active use of a natural environment and landscape
	- Expansion and development of local communities
	- Creation of a foundation for strengthening the local community capacity
	- Possible contribution to local economic revitalization
Section	Success Factors	Policies Support
Topic 1: Local cultural experience	- Local cultural - resources - Local arts resources - Experience - programs	- Landscape character assessment, local amenity survey - Training of local cultural commentators/> - Education and support for the inheritance of traditional culture
Topic 2: Utilization of environment and landscape	- Conservation of - environment and landscape - Ecotourism	- Environment-friendly agriculture and environment–village certification system - Establishing ecotourism infrastructure
Topic 3: Community Activation	- Citizen participation - Cooperation - Community building	- Introduction of a participatory system for establishing local plans - Establishment of exchange programs between returning farmers and existing residents
Topic 4: Establishment of local infrastructure	- Expansion of living and economic infrastructure - Improvement of - living conditions	- Establishing a digital rural platform - Supporting small-scale maintenance-type living SOC - Expanding social infrastructure and improving accessibility in each village
Topic 5: Local economic activation	- Strengthening rural self-reliance - Market expansion - Local branding	- Ruralism brand development - Support for developing direct transaction platforms and online and offline sales channels

summarizes the five topics derived through LDA topic modeling and the corresponding success factors associated with each. This study is significant in that it is one of the first to explore the practical directions of rural spatial policy.

5.1. Success Factor Application and Regional Adaptation Strategies

As the relative importance of each success factor may vary depending on regional characteristics, it is necessary to adopt a differentiated weighting strategy in policy design. Rural areas exhibit diverse geographical forms, such as mountainous, coastal, and plain terrains, and show considerable differences in resident composition and socioeconomic conditions. These disparities imply that a uniform policy tool does not exert uniform effects across all regions. Accordingly, it is essential to adjust the weight of each success factor to reflect local contexts.

A particularly noteworthy point is that many rural villages in reality are not confined to a single thematic focus but rather exhibit multiple characteristics and functions, so-called multi-theme villages. For example, a village promoting rural tourism may simultaneously experience revitalization in both community engagement and the local economy. In this study, one such example is a village in Hapcheon-gun, Gyeongsangnam-do, which simultaneously reflects characteristics of both Topic 1: local cultural experience and Topic 5: local economic activation, demonstrating a tendency toward multi-themed revitalization. Accordingly, in the design and implementation of rural policies, a multilayered analytical framework and integrated response strategies are required to account for the interactions among different themes and the complex nature of success factors [64].

Specific application methods may include setting policy priorities based on resident demand surveys and employing policy simulation tools. Resident demand surveys involve evaluating the perceived importance of each success factor through questionnaires and focus group interviews with residents. Based on these results, weights can be assigned to calculate a policy priority index. Currently, rural policy in Korea is largely based on a top-down approach led by the central government. However, as residents possess the most accurate understanding of their region’s specific needs and characteristics, a shift toward a resident-led governance model is required, like the EU’s LEADER program [65,66]. Capturing resident needs also enhances the legitimacy, acceptance, and sustainability of rural development policies as part of a bottom-up strategy [67,68].

Simulation of policies serves as an infrastructure for forecasting the impacts of proposed policies. Such tools are increasingly used to assess the effectiveness of legal reforms and to evaluate whether intended outcomes are likely to be achieved [69]. Notably, EUROMOD has been developed to assist EU member states in evaluating the impact of tax and welfare policies. These simulations enable analyses of the effects of existing policy changes on income inequality and poverty and can project future scenarios involving hypothetical changes in policy, economy, or demography [70]. Though primarily used in applied economics, policy simulation has potential applications in rural planning. For instance, before allocating rural revitalization subsidies, simulations can be conducted to understand the regional consumption patterns and income levels, enabling more efficient and tailored support. Likewise, in the context of policies supporting the return of young people to rural areas, simulations can be utilized to estimate education and housing demands, as well as the broader regional economic impacts.

In the case of multi-theme villages that encompass multiple themes simultaneously, it is necessary to prioritize elements that are commonly integrated across themes. Elements such as resident participation, utilization of the region’s unique environment and cultural assets, and contributions to the local economy are relevant across all five thematic categories. Therefore, by prioritizing these common elements in the design and implementation of policies, it is possible to maximize the synergistic effects among themes and enhance both the coherence and sustainability of rural development strategies.

Integrating the success factors identified in this study into the rural policy-making process can significantly enhance the relevance and effectiveness of planning and policy design. Given the environmental and socioeconomic heterogeneity of rural areas, uniform application of policy criteria may be limited in effectiveness. Therefore, systematic discussions to establish universally acceptable policy frameworks and methods for region-specific implementation are indispensable.

5.2. Comparison with International Examples of Rural Revitalization

Although this study focuses on rural areas in South Korea, issues such as rural population decline and aging are global. Therefore, it is necessary to consider rural revitalization policies from an international perspective [71]. In many countries, major plans and policies related to rural revitalization are being implemented in diverse ways. Japan, the UK, and France have established strategies that reflect the specific characteristics and social structures of their respective regions.

Japan is one of the countries that began implementing rural revitalization policies relatively early [72]. According to the Basic Plan for Food, Agriculture and Rural Areas, rural development is structured around three main pillars: comprehensive rural development, development of mountainous areas, and urban–rural exchange. Among these, comprehensive rural development focuses on strengthening the inherent functions and roles of rural areas, independent of agricultural policy. In particular, the aging of rural populations, which represents one of the most serious challenges facing Japan’s rural areas, has been recognized as a critical policy issue. Enhancing the self-sufficiency of rural spaces has been set as a policy priority. Furthermore, the approach emphasizes not only agricultural support but also integrated strategies encompassing social, economic, and environmental development [73].

The United Kingdom has proposed four key priorities for rural revitalization: economic growth, connectivity, housing and energy, and community empowerment. Policies such as the Countryside Protection Act and the Rural Development Programme have been introduced, most of which focus on preserving cultural and ecological values. These initiatives aim to protect the natural environment and landscape characteristics of each region while promoting the development of economic activities that reflect local identities [74].

Meanwhile, France initially implemented rural support policies based on population size [22]. Still, it later expanded its approach through the public research program known as PSDR (Pour et Sur le Développement Régional) as part of its rural revitalization strategy [75]. The PSDR program conducts region-specific research centered on the five major themes: farmland governance, landscape preservation and quality of life, territorial attractiveness, territorialized food systems, and the circular economy. It places particular emphasis on the development of pilot tools for implementing actual policies [76]. Among these themes, territorial attractiveness has emerged as a key research focus. The program aims to move beyond the expansion of basic physical infrastructure, such as through conventional infrastructure and innovation policies, and instead pursue sustainable and differentiated rural development models through the strategic utilization of local resources and identities [77,78].

These international cases reflect a global consensus on the importance of formulating strategies that consider the unique characteristics and local contexts of rural areas. They offer meaningful insights into the future direction of rural revitalization policies [71,79,80,81].

5.3. Limitations and Future Research

This study underscores the importance of closely examining and seeking practical measures for the tangible and intangible resources that can be utilized in each region and potentially lead to external engagement. The five suggested types—local cultural experience, utilization of environment and landscape, community activation, establishment of local infrastructure, and local economic activation—offer fresh perspectives as discussion points applicable across diverse rural settings.

Nevertheless, several limitations should be acknowledged. First, the dataset was constructed based on successful cases of rural revitalization projects. While this approach is useful for extracting key success factors, it may entail selection bias, which could limit the generalizability of the findings. Future research would benefit from including a broader dataset that encompasses a wider range of rural development outcomes, including less successful or failed cases.

Second, the analysis employed the traditional TF-IDF algorithm for feature extraction. Although effective, this method may fall short when handling long-tail or contextually nuanced terms. Incorporating advanced word embedding techniques such as Word2Vec or FastText in future studies could enhance the robustness of topic modeling results.

Despite these limitations, the study provides a meaningful contribution by identifying key success factors through topic modeling based on exemplary rural cases and offering policy guidelines for rural revitalization. The addition of field-based empirical research in future studies would further validate the findings and enhance their applicability.

6. Conclusions

As socioeconomic development following industrialization is mainly centered on urban areas, well-established rural revitalization policies and plans are needed to prevent regional extinction and revitalize them [82]. However, the rural revitalization projects promoted thus far lack consideration of the specific characteristics of each region, and many have yielded only short-term, visible results. Ultimately, plans and policies that consider regional characteristics and differences are necessary to achieve the expected regional revitalization. This study aimed to provide core guidelines that can be universally utilized when establishing rural plans and policies. In addition, by analyzing the socioeconomic outcomes resulting from regional development programs based on objective data, the goal was to provide a direction for rural plans in a practical rather than symbolic sense.

A decade of massive rural revitalization projects in South Korea was analyzed, and text-mining techniques were used to ensure the objectivity and diversity of interpretation of the results. Specifically, after evaluating the appropriateness of the number of clusters through k-means clustering analysis and deriving five clusters, LDA topic modeling was performed to derive topics affecting the success of rural revitalization projects.

The five topics derived from this study can serve as practical and essential components to be considered in the formulation of rural revitalization plans and policies. By applying differentiated weights to each factor according to the unique characteristics and conditions of individual regions, it becomes possible to establish more precise and effective tailored strategies. This classification framework can serve as a basis for the development of strategic policies that account for interregional differences and contribute to the design of sustainable and differentiated rural development strategies.