2020 Volume 76 Issue 2 Pages I_20-I_31
From the 1930s to the present, the ratio of construction industry to fatalities in all industries has been high and the need for countermeasures is high. One of the countermeasures is the use of case studies categorized by the nature of the work, but the number of categorized case studies is very small and does not lead to sufficient learning.
In this study, the classification tendency and classification method of the work contents were examined by using the document data of the disaster situation common to all the case studies, paying attention to the quantitative text analysis as a method to classify the existing construction accident case studies by the work contents.
As a result, by checking the co-occurrence of parts of speech related to actions and materials and work locations in the entire civil engineering industry data, we were able to extract work content relevant to the main types of accidents. A classification method focusing on co-occurrence was examined, and a classification method by work content using quantitative text analysis was organized for the major cases in the case collection.