| Title |
Characteristics Analysis of Seasonal Construction Site Fall Accident using Text Mining |
| Authors |
Kim, Joon-Soo ; Kim, Byung-Soo |
| DOI |
http://dx.doi.org/10.6106/KJCEM.2019.20.3.113 |
| Keywords |
Industrial Accidents per 100 ; 000 People ; Cluster Analysis ; Principal Component Analysis ; Falling Accidents ; Text Mining |
| Abstract |
The death rate of industrial accidents per 10,000 people in Korea is two to three times higher than that of major countries. Falling accidents at the construction site happened to have caused the most deaths. Analysis of existing research and measures by national institutions showed that the industrial accident management concentrated on falling accidents was insufficient and the seasonal safety management measures were not enough. There is thus the need for research that provides detailed and enough information on falling accidents. This study, therefore, aims to overcome the limitations of existing research and safety management accident response using a methodology that provides the necessary information for the prevention of fall accidents by deriving seasonal crash characteristics of the construction site. In order to provide enough information, 387 cases of seasonal construction site falling were collected, which describes the causal relationship of accidents. Text mining using principal component analysis and cluster analysis was carried out. The analysis showed that: In the spring, snowfall and unreasonable operation of equipment including lifts were the major cause. In summer, most accidents were caused by form, insufficient safety inspection, and installation work. In autumn, weather factors such as wind and typhoon were the cause. In winter, material transportation, exterior wall work, and ignore safety precautions were the cause of the crash. |