합성곱신경망을 이용한 안전문화 수준 분류

다국어 초록 (Multilingual Abstract)

Objective: The aim of this study is to investigate whether an artificial intelligence technique can be successfully applied to safety culture survey data classification.

Background: Without help of artificial intelligence technique, classifying safety culture level from safety culture scores collected from a large-scale survey would require a lot of experts" time and effort.

Method: Two convolution layers and 1 pooling layer was used as the middle layer of the artificial neural network to design a classifier for safety culture level. 1045 safety culture survey data collected from power plant workers used to train and validate the classifier.

Results: After 40 epochs" training the classifier approached near 95% precision in classifying "safe" and "need to be improved" classes of safety culture level. Precision, recall and F1-score for the test data set showed over 95% of accuracy performance.

Conclusion: An artificial intelligence technique using such as convolution neural network can help classification of safety culture survey data.

Application: The safety culture level classifier using deeper neural network and big survey data might improve the performance and might substitute expert interviewers for safety culture evaluation.

목차 (Table of Contents)

• 1. Introduction
• 2. Method
• 3. Results and Conclusions
• References

참고문헌 (Reference)

1 이동하, "Suggestions for More Reliable Measurement of Korean Nuclear Power Industry Safety Culture" 대한인간공학회 35 (35): 75-84, 2016

2 Muniz, B. F., "Safety culture: analysis and the causal relationships between its key dimensions" 38 : 627-641, 2007

3 이용희, "Safety Culture, A New Challenge to Human Factors Engineering for 21st Century" 대한인간공학회 35 (35): 473-492, 2016

4 Olive, C., "Relationship of safety culture and process safety" 130 : 133-140, 2006

5 Hale, A., "Occupational Injury: Risk Prevention, and Intervention" Taylor & Francis 129-166, 1998

6 김대호, "Measuring Safety Culture to Promote Aviation Safety Culture" 대한인간공학회 35 (35): 111-123, 2016

7 Reason, J., "Managing the Risks of Organizational Accidents" Ashgate 1997

8 Stanton, N. A., "Human Factors in the Design and Evaluation of Central Control Room Operations" CRC Press 2010

9 Geron, A., "Hands-On Machine Learning with Scikit-Learn and TensorFlow: Concepts, Tools, and Techniques to Build Intelligent Systems" O'Leilly 2017

10 이용희, "Current Status and Issues of Nuclear Safety Culture" 대한인간공학회 35 (35): 247-261, 2016

11 Pidgeon, N. F, "Aviation Psychology in Practice" Aldershot 21-43, 1994

12 Ball, P. W., "Approaches to Safety Culture Enhancement" British Nuclear Fuels, Ltd 1998

13 Health and Safety Executive (HSE), "A Review of the Safety Culture and Safety Climate Literature for the Development of the Safety Culture Toolkit" HSE Books 2005

14 김범수, "A Methodology for Establishment of Safety Culture in Industry through Global Company Practices" 대한인간공학회 35 (35): 85-95, 2016