이차희 Chahee Lee , 김용진 Youngjin Kim

DOI:10.7851/ksrp.2026.32.3.001 Vol.32(No.3) 1-11, 2026

---

Recent population decline in Korea’s rural areas is not simply a matter of fewer residents, but also a matter of reshaping the spatial and functional conditions under which rural areas provide services and pursue endogenous development. To this end, this study examines rural-urban linkages as a way to reconsider the roles rural areas can play in relation to cities. Drawing on neo-endogenous development theory and OECD debates on low-density economies and rural-urban linkages, it analyzes survey data from 304 local government officials and rural policy experts in Korea, and found that there is a certain gap between existing linkage practices and the future directions perceived by policymakers. The results show that current rural-urban linkages are skewed toward in specific sectors such as tourism, experiential activities, general regional development, and agricultural and livestock product trade. These linkages tend to be limited to within individual local government boundaries and do not extended beyond their boundaries; the entities participating in these linkages are limited, with businesses and institution from new sectors such as technologies and energy difficult to join. Future demand for linkages was, however, anticipated to expand to include new functions such as flexible residency, telemedicine and remote education, and new technology demonstration. These findings suggest that rural-urban linkage policies need to expand beyond the existing framework of urban-rural exchange into diverse functional areas, and that it is necessary to establish a governance system and institutional foundation to support this.

김해영 Kim Hae-young , 조중현 Cho Joong-hyun

DOI:10.7851/ksrp.2026.32.3.013 Vol.32(No.3) 13-28, 2026

---

This study selected Hongnong-eup, Yeonggwang-gun, Jeollanam-do as a case area and applied the KAP (Knowledge-Attitude-Practice) model to analyze the correlation and practice influencing factors of residents’ knowledge, attitude, and practice in landscape management, and conducted a correlation analysis including not only the linear flow of ‘knowledge → attitude → practice’ but also interaction and cyclical relationships. As a result, attitude acted as the most important influencing factor in practice, and the correlation between ‘problem recognition’, ‘individual landscape management activity experience’, and ‘community participation experience’ was high as detailed influencing factors. In addition, results were derived indicating that landscape management performance may vary depending on whether or not the community is activated. The results of the study revealed that the landscape management policy in rural area development projects should be designed in a ‘participation-learning type’ and ‘community activation’ structure that induces a change of perception through practice, away from simple knowledge and information delivery-oriented education, and the administration suggested that follow-up support considering sustainability, such as fostering leaders, evaluating residents’ participation, and establishing a network of experts, is necessary so that the residents’ community can continue to operate.

최상원 Choi Sangwon , 서동주 Suh Tongjoo

DOI:10.7851/ksrp.2026.32.3.029 Vol.32(No.3) 29-40, 2026

---

박근오 Park Keun-oh , 이춘호 Lee Chun-ho

DOI:10.7851/ksrp.2026.32.3.041 Vol.32(No.3) 41-56, 2026

---

The Rural Space Restructuring and Regeneration Support Act (2024) established seven specialized districts to prevent rural sprawl and promote regional specialization; this study addresses the pre-amendment seven-district system (i.e., prior to the November 2025 enforcement decree amendment that added an eighth district). Existing designation methods rely on simple area ratios or subjective judgment. This study classifies building gross floor area from 1,202 non-capital-region townships into the seven district categories. Because the Entropy and HHI indices are arithmetic functions of the same seven floor-area shares used as independent variables, results involving these two indices are interpreted as descriptive/compositional associations rather than causal effects; only the population growth model is treated as an independent dependent variable. OLS, SLM, and SEM models and Moran’s I / LISA analyses are used as exploratory tools. Results show PROTECT and SCENIC floor-area shares are associated with higher Entropy, while LIVESTOCK share is associated with lower population growth and Entropy and higher HHI; explanatory power for population growth remains low (R² ≈ 0.07-0.08) even after spatial modeling, indicating that floor-area alone cannot account for most population dynamics. LISA hotspots are presented as preliminary, exploratory reference information rather than direct designation criteria. This study proposes a three-step preliminary diagnostic procedure - floor-area classification, spatial econometric modeling, and LISA pattern exploration - as a foundation for future work incorporating non-building spatial data.

강동현 Kang Dong-hyeon , 최영완 Choi Young-wan , 김영주 Kim Young-joo

DOI:10.7851/ksrp.2026.32.3.057 Vol.32(No.3) 57-65, 2026

---

This study classified eup-myeon areas using Davies’s Functional Index and the Gini coefficient of welfare facility distribution to examine improvement strategies for Rural Development Projects, addressing the limitations of the current classification system based primarily on central-place hierarchy. The study focused on Hadong-gun and Hapcheon-gun in Gyeongsangnam-do, Korea. Centrality was analyzed using seven Living SOC facility categories―education, welfare, culture, sports, health, commerce, and administration―while the Gini coefficient was calculated using population and welfare facility data at the legal village (Ri) level. The results showed a weak relationship between the Functional Index and the Gini coefficient, indicating that central-place hierarchy does not necessarily correspond to welfare facility distribution within eup-myeon areas. Furthermore, different levels of welfare facility distribution were observed within the same centrality class. These findings suggest that centrality alone is insufficient to explain variations in welfare facility distribution and highlight the need to consider both centrality and welfare facility distribution when classifying project types and developing improvement strategies for Rural Development Projects.

이지민 Lee Jimin

DOI:10.7851/ksrp.2026.32.3.067 Vol.32(No.3) 67-77, 2026

---

This study explores the use of AutoEncoder-based representation learning for classifying rural grid attributes in Hongcheon County, South Korea. The population, number of buildings, slope, and road length were calculated at the 500-meter grid level and compressed into a two-dimensional embedding using an AutoEncoder. HDBSCAN and K-Means were then applied to classify grids with similar attributes. The AutoEncoder achieved an overall reconstruction R² of 0.969 on the held-out validation data, with variable-specific R² values ranging from 0.951 to 0.984. HDBSCAN identified four clusters and classified 15.5% of the grids as unclassified observations (noise). The silhouette coefficient excluding these grids was 0.543, and the DBCV was 0.025. K-Means classified all grids into five clusters with a silhouette coefficient of 0.480, slightly higher than that of the PCA-based representation (0.455) and higher than that of the original standardized data (0.352) under the same conditions. External validation using daily-life service facilities showed significant differences in several facility variables across clusters. However, cluster assignments varied across repeated AutoEncoder training runs. Therefore, the derived groups should be interpreted as exploratory classifications of rural grid attributes rather than fixed spatial hierarchies or actual living-area boundaries. This approach may be useful for exploratory rural spatial analysis and could provide a basis for future research incorporating accessibility, mobility, and service-use data.

안옥선 Ahn Ok-sun , 조숙영 Cho Suk-yeong , 서기춘 Seo Ki-chun , 조한솔 Cho Han-sol , 김상범 Kim Sang-bum , 배승종 Bae Seung-jong

DOI:10.7851/ksrp.2026.32.3.079 Vol.32(No.3) 79-92, 2026

---

This study examines implementation challenges perceived by municipal officials responsible for preparing city/county rural spatial master plans during the early implementation period of Korea’s Rural Spatial Restructuring and Regeneration Support Act. A structured survey was administered from September 10 to 19, 2025, to officials in municipalities that had completed or were preparing rural spatial plans, and valid responses from 48 officials in 42 municipalities were analyzed. The analysis combined frequency analysis, descriptive statistics, multiple-response analysis, and Mann-Whitney U tests. The results show that the governance system required for rural spatial planning has been only partially institutionalized. Most municipalities relied on existing departments rather than newly established dedicated administrative units, and shortages of personnel and expertise were the most salient barriers to the effective operation of governance bodies. In the decision-making process, financial and budget feasibility, resident demand, and compliance with laws and guidelines were the main criteria, whereas lack of interdepartmental cooperation, insufficient staff expertise, and unclear decision-making authority were the primary difficulties. Regarding information acquisition, officials reported the greatest difficulties with data on rural hazardous facilities, rural environmental management facilities, environmental and landscape preservation, and rural communities; they particularly emphasized problems of data currency, fragmented data platforms, and data reliability. The findings suggest that early stabilization of the rural spatial planning system requires an integrated strategy linking governance capacity, interdepartmental coordination, and rural spatial information infrastructure. Policy priorities include supporting the effective operation of dedicated administrative units and administrative councils, clarifying the roles of local support agencies, deliberation committees, and resident consultative bodies, strengthening professional training, and developing a GIS-based rural spatial information platform that supports evidence-based designation and management of rural regeneration activation areas and rural specialized districts.

김종필 Kim Jong-pil , 김태연 Kim Tae-yeon , 윤상헌 Yun Sang-hun , 임청룡 Lim Cheong-ryong

DOI:10.7851/ksrp.2026.32.3.093 Vol.32(No.3) 93-100, 2026

---

This study analyzes differences in net returns and cost structures across paddy rice farm-size groups in Korea. Using a log-linear regression model with interaction terms, the study examines whether the relationships between material, contract farming, and labor costs and net returns differ among small-, medium-, and large-scale groups. The estimation results show that medium- and large-scale groups have higher conditional levels of net returns than the small-scale group, while the coefficients of the cost variables also differ across farm-size groups. These results indicate heterogeneity in management performance and cost structures by farm size. However, the estimated differences should not be interpreted as causal effects of farm expansion or as direct evidence of economies of scale. In addition, because net returns are mechanically related to production-cost components, the cost coefficients are interpreted as conditional associations rather than independent productivity effects. The findings suggest that agricultural policies should consider differences in management conditions and cost burdens across farm-size groups.

오윤정 Oh Yoonjung , 김혜리 Jin Huili , 안동환 An Donghwan

DOI:10.7851/ksrp.2026.32.3.101 Vol.32(No.3) 101-114, 2026

---

This study examines how North Korea’s rural development strategy under the Kim Jong-un regime is reflected in the internal spatial structure of rural counties. Orang County in North Hamgyong Province and Sukchon County in South Pyongan Province, where both the Agricultural Development Zone initiative and the Regional Development 20×10 Policy have been implemented, were selected as case studies. The analysis combines GIS-based spatial analysis with physiographic, land-use, and socioeconomic data, supplemented by documentary evidence from North Korean publications, to investigate the relationship between the location of development projects and county-level spatial structure. The findings show that the two counties exhibit distinct internal spatial structures while sharing a common pattern in which local economies are organized around established agricultural production bases and existing county centers. Both development initiatives reinforce existing production systems and spatial structures by locating development projects in potential development areas adjacent to established development cores, rather than creating entirely new growth centers. These findings suggest that North Korea’s local development policy strengthens county-level production, processing, and distribution systems by building upon the hierarchical spatial structure and functional linkages within counties. By providing a micro-scale analysis of county-level spatial structures, this study offers empirical evidence on the spatial logic underlying North Korea’s rural development strategy and provides a foundation for evaluating future regional development policies and identifying potential areas for inter-Korean cooperation.

박엄용 Park Oumyong , 리재웅 Rhee Zaewoong , 박지현 Park Ji-hyun , 김수진 Kim Soo-jin

DOI:10.7851/ksrp.2026.32.3.115 Vol.32(No.3) 115-132, 2026

---

This study examines the relationship between AI literacy and job continuity among employees of intermediary support organizations, focusing on the mediating role of job efficacy and the moderating role of job autonomy. Data were collected through an online survey of employees working in intermediary support organizations across South Korea, and 217 valid responses were analyzed. Confirmatory factor analysis was conducted to assess the validity and reliability of the measurement model, followed by regression-based mediation and moderation analyses using the SPSS PROCESS Macro. Gender, age, and organizational tenure were included as control variables. The results showed that AI literacy had a significant positive effect on job efficacy and a significant positive effect on job continuity. Job efficacy also had a significant positive effect on job continuity. Furthermore, job efficacy partially mediated the relationship between AI literacy and job continuity, indicating that employees with higher AI literacy were more likely to perceive themselves as capable of performing their jobs effectively, which in turn enhanced their job continuity. Job autonomy significantly moderated the relationship between AI literacy and job continuity. Specifically, the positive relationship between AI literacy and job continuity became stronger as job autonomy increased. These findings demonstrate that AI literacy functions as an important individual resource that can enhance employees’ confidence in performing their jobs and promote workforce stability. The study suggests that intermediary support organizations should provide job-related AI literacy training while creating an organizational environment in which employees can autonomously apply AI tools and competencies to their work.