DOI QR코드

DOI QR Code

Prediction Model of Construction Safety Accidents using Decision Tree Technique

의사결정나무기법을 이용한 건설재해 사전 예측모델 개발

  • Cho, Yerim (Department of Architectural Engineering, Graduate School, Kyonggi University) ;
  • Kim, Yeon-Choel (Graduate School of Construction Industry, Kyonggi University) ;
  • Shin, Yoonseok (Department of Plant.Architectural Engineering, Kyonggi University)
  • Received : 2017.01.31
  • Accepted : 2017.03.27
  • Published : 2017.06.20

Abstract

Over the past 7 years, the number of victims of construction disasters has been gradually increasing. Compared with projects in other industries, construction projects are highly exposed to safety risks. For this reason, the research methods of predicting and managing the risk of construction disasters are urgently needed that can be applied to a construction site. This study aims to propose a prediction model for a construction disaster using the decision tree technique. The developed the model is reviewed the applicability by evaluating its accuracy based on disaster data. The top three of the prediction values obtained from the proposed model were enumerated, and then the cumulative accuracy were also calculated. The prediction accuracy was 40 percent for the first value, but the cumulative accuracy was 80 percent. Thus, as more disaster data was accumulated, the cumulative accuracy appeared to be higher. If utilized in construction sites, the model proposed in this study would contribute to a reduction in the rate of construction disasters.

건설 산업 재해 예방을 위한 연구와 노력에도 불구하고 최근 7년간 국내 건설업 재해자 수가 꾸준히 증가했다. 건설현장에서 발생하는 재해는 다른 산업군에 비해 강도 높은 재해가 발생할 가능성이 크기 때문에 근본적으로 예방할 수 있는 방법이 필요하다. 따라서 본 연구에서는 모형에 대한 해석이 쉽고 변수의 상호작용 효과 해석이 용이한 의사결정나무 기법을 활용하여 건설재해 예측 모델을 제안하였다. 제안된 건설 재해 사전 예측 모델의 현장 활용 가능성을 평가하기 위하여 판별분석기법 기반 모델과의 건설 재해 예측 정확도를 비교하였다. 검토 결과 판별분석 모델에 비해 의사결정나무 모델의 누적 예측 정확도가 더 높은 것으로 나타났다. 의사결정나무 기법을 이용한 모델은 시간이 지남에 따라 데이터가 증가하기 때문에 예측 정확도가 더욱 높아지게 된다. 따라서 본 연구에서 제안된 건설 재해 예측 모델이 건설현장에서 활용된다면 효과적으로 안전 관리를 할 수 있고, 건설업 재해율 감소에도 기여할 수 있을 것으로 기대한다.

Keywords

References

  1. Ministry of Employment and Labor. 2010 Industrial accident analysis of current situation. Sejong-si (Korea): Ministry of Employment and Labor; 2011. 454 p.
  2. Ministry of Employment and Labor. 2011 Industrial accident analysis of current situation. Sejong-si (Korea): Ministry of Employment and Labor; 2012. 439 p.
  3. Ministry of Employment and Labor. 2013 Industrial accident analysis of current situation. Sejong-si (Korea): Ministry of Employment and Labor; 2014. 529 p.
  4. Ministry of Employment and Labor. 2015 Industrial accident analysis of current situation. Sejong-si (Korea): Ministry of Employment and Labor; 2016. 759 p.
  5. An SH, Ryu HW, Park MS. Selecting Key Accident Risk Factors in Building Construction Projects. Journal of Korea Institute of Building Construction. 2011 Feb;11(1):45-50. https://doi.org/10.5345/JKIC.2011.02.1.045
  6. Shin YS, Yoo WS. Early Warning Model using Case-Based Reasoning for Construction Site Safety Accidents. Journal of Korean society of hazard mitigation. 2015 Dec;15(6):27-33. https://doi.org/10.9798/KOSHAM.2015.15.6.27
  7. Lee DG, Lee DE. Data mining in construction fatal accident: association rule mining. Proceeding of the Association of Architectural Institute of Korea; 2010 Dec 4; Busan, Korea. Busan (Korea): the Association of Architectural Institute of Korea; 2010. p. 426-30.
  8. Kim JH. The improvement of providing protective equipment preventing a fall in small scale construction site [master's thesis]. [Suwon (Korea)]: Kyunghee University; 2013. 76 p.
  9. Kim TH. Advanced system formwork and construction planning model for tall building construction [dissertation]. [Seoul (Korea)]: Korea University; 2013. 140 p.
  10. Lee J. A study of churn prediction model for long-term customers using data mining technique [master's thesis]. [Seoul (Korea)]: Korea University; 2013. 63 p.
  11. Kim JU, Kim ST. Data mining techniques for marketing, sales, and customer relationship management. 3rd ed. Seoul (Korea): Hankyungsa; 2012. 627 p.
  12. Jeong GY. Multivariate control chart procedure using decision tree learning technique [master's thesis]. [Seoul (Korea)]: Chung-Ang University; 2015. 36 p.
  13. Yoo SD. A study on the stability of non-response imputation using CHAID Algorithm. [master's thesis]. [Seoul (Korea)]: Korea University; 2015. 33 p.
  14. Leem YM, Hwang YS. A Feature Analysis of Industrial Accidents Using CHAID Algorithm. Journal of Korea Safety Management & Science. 2005 Dec;7(5):59-67.
  15. Park JJ, Seo IK, Kang DY, Lee JH. A Safety Evaluation of Shoulder Rumble Strips on Expressway using Discriminant Analysis. Journal of the Korean Society of Safety. 2016 Feb;31(1):95-104. https://doi.org/10.14346/JKOSOS.2016.31.1.095
  16. Kim YC, Yoo WS, Shin YS. Application of artificial neural networks to prediction of construction safety accidents. Journal of Korean society of hazard mitigation. 2017 Feb;17(1):7-14. https://doi.org/10.9798/KOSHAM.2017.17.1.7

Cited by

  1. Factors related to the Identification of Middle-Aged Women Who are Disadvantaged by Cardio-cerebrovascular Disease vol.24, pp.2, 2018, https://doi.org/10.4069/kjwhn.2018.24.2.185