All Issue

2026 Vol.22, Issue 3 Preview Page

Original Article

Analysis of Factors Associated with CCTV Installation in Child Protection Zones Using Machine Learning and Empirical Data Refinement

머신러닝과 실증적 데이터 정제 기법을 활용한 어린이보호구역 CCTV 설치 요인 분석

구창민1ㆍ고승욱2*

Changmin Gu1, Seungwook Go2*

1Associate Research Fellow, Disaster & Safety Center, Chungnam Institute, Gongju, Republic of Korea
2Associate Research Fellow, AI & Data Research Team, Chungnam Institute, Gongju, Republic of Korea
30 September 2026. pp. 800-815
PDF Citation
Abstract
Purpose: This study aims to identify CCTV non-installation records in child protection zones and explore factors associated with CCTV installation. Method: The 2026 National Standard Data on Child Protection Zones were refined by correcting inconsistencies between CCTV installation status and unit counts. Descriptive analysis and municipality-level GIS mapping were conducted on 14,635 records, and a random forest model, Welch’s t-test, and sensitivity analyses were applied to 10,572 valid observations. Result: A total of 3,133 zones (21.4%) were recorded as CCTV non-installed. The random forest model showed an OOB accuracy of 83.92% and an AUC of 0.784. Metropolitan/provincial region, road width, and facility type were identified as key associated variables, and the results remained consistent across sensitivity analyses. The non-installed group had a significantly narrower mean road width. Conclusion: CCTV non-installation records are associated with regional heterogeneity, facility type, and physical road conditions. Policy support should consider standard-data quality management, road-space constraints, and facility-specific infrastructure needs.
Keywords
Child Protection Zone
CCTV
Safety Infrastructure
Random Forest
Road Width
연구목적: 본 연구는 전국 어린이보호구역의 CCTV 미설치 기록을 파악하고, CCTV 설치 여부와 관련된 요인을 탐색하는 데 목적이 있다. 연구방법: 행정안전부의 2026년 전국 어린이보호구역 표준데이터를 활용하여 설치 여부와 설치 대수 간 불일치 보정 등 정제 과정을 수행하였다. 전체 14,635개소에 대한 기술통계와 시군구 단위 GIS 분포 분석을 실시하고, 유효 10,572개소를 대상으로 랜덤 포레스트 분석, Welch의 t-검정, 민감도 분석을 수행하였다. 연구결과: 전체의 21.4%인 3,133개소가 공식 데이터상 CCTV 미설치로 분류되었다. 랜덤 포레스트 모형의 OOB 정확도는 83.92%, AUC는 0.784였으며, 광역시도, 도로폭, 시설종류가 주요 관련 변수로 확인되었고 이 결과는 민감도 분석에서도 유지되었다. 미설치 그룹은 설치 그룹보다 평균 도로폭이 유의하게 좁았다. 결론: 어린이보호구역의 CCTV 미설치 기록은 지역적 이질성, 시설 유형, 도로 환경과 관련될 수 있다. 향후 정책은 표준데이터 품질 관리, 협소 도로의 설치 제약, 시설 유형별 인프라 수요를 함께 고려할 필요가 있다.
키워드
어린이보호구역
CCTV
안전 인프라
랜덤 포레스트
도로폭
References
  1. Breiman, L. (2001). “Random forests.” Machine Learning, Vol. 45, No. 1, pp. 5-32. 10.1023/A:1010933404324
  2. Chung, S., Lee, J., Kim, Y., Lee, C. (2023). “Analysis of influence area of speed camera in school zone based on survival analysis.” Journal of Korean Society of Transportation, Vol. 41, No. 4, pp. 425-439. 10.7470/jkst.2023.41.4.425
  3. Eun, S.J. (2023). “Effects of tougher school zone laws on road traffic safety in school zones for children in South Korea.” Journal of Transport & Health, Vol. 32, Article 101687. 10.1016/j.jth.2023.101687
  4. Joo, B.-C. (2023). “Status and improvement plans for sidewalk safety facilities (barrier) in child protection area.” Journal of Korean Society of Disaster & Security, Vol. 16, No. 4, pp. 137-145. 10.21729/KSDS.2023.16.4.137
  5. Korea Artificial Intelligence Hub (2022). Video Data of Children’s Risky Road-Walking Behaviors in Child Protection Zones. Seoul, Korea.
  6. Liaw, A., Wiener, M. (2002). “Classification and regression by randomforest.” R News, Vol. 2, No. 3, pp. 18-22.
  7. Lim, S., Choi, J., Oah, S. (2020). “Effects of gateway on reducing vehicle speeds in school zones.” Journal of Korean Society of Transportation, Vol. 38, No. 2, pp. 85-96. 10.7470/jkst.2020.38.2.085
  8. Ministry of the Interior and Safety (2026a). National Standard Data on Child Protection Zones. Public Data Portal, Sejong, Korea.
  9. Ministry of the Interior and Safety (2026b). 2026 Child Safety Implementation Plan Confirmed, Focusing on Eight Safety Management Areas. Sejong, Korea.
  10. National Law Information Center (2026). Road Traffic Act, Article 12: Designation, Revocation, And Management of Child Protection Zones. Ministry of Government Legislation, Sejong, Korea.
  11. Park, S.-W., Yoo, S.-M., Kim, G.-W. (2023). “Trends in the application of intelligent cctv technology in public safety.” Weekly ICT Trends, Vol. 2105, pp. 16-28.
  12. Piza, E.L., Welsh, B.C., Farrington, D.P., Thomas, A.L. (2019). “CCTV surveillance for crime prevention: A 40-year systematic review with meta-analysis.” Criminology & Public Policy, Vol. 18, No. 1, pp. 135-159. 10.1111/1745-9133.12419
  13. Robin, X., Turck, N., Hainard, A., Tiberti, N., Lisacek, F., Sanchez, J.-C., Müller, M. (2011). “pROC: An open-source package for R and S+ to analyze and compare ROC curves.” BMC Bioinformatics, Vol. 12, Article 77. 10.1186/1471-2105-12-77 21414208 PMC3068975
  14. Strobl, C., Boulesteix, A.-L., Zeileis, A., Hothorn, T. (2007). “Bias in random forest variable importance measures: illustrations, sources and a solution.” BMC Bioinformatics, Vol. 8, Article 25. 10.1186/1471-2105-8-25 17254353 PMC1796903
Information
  • Publisher :The Korean Society of Disaster Information
  • Publisher(Ko) :한국재난정보학회
  • Journal Title :Journal of the Society of Disaster Information
  • Journal Title(Ko) :한국재난정보학회논문집
  • Volume : 22
  • No :3
  • Pages :800-815
  • DOI :https://doi.org/10.15683/kosdi.2026.9.30.800
Journal Informaiton Journal of the Society of Disaster Information Journal of the Society of Disaster Information
  • NRF
  • KISTI Current Status
  • KISTI Cited-by
  • crosscheck
  • orcid
  • open access
  • ccl
Journal Informaiton Journal Informaiton - close