A Study on the Improvement of Classification of Road Problem Reporting Types

In this study, a total of 7,448 reports occurred on the road were analyzed, and keywords that can represent the report type were analyzed and presented through the report contents.Keywords such as road signs, barrier, traffic lights, and guard-rails were typically derived for road facilities, and keywords indicating conditions such as cracks, and irregularities were derived for road surface conditions.Road kill/rock fall was derived from the conditions of animals frequently found on roads such as animal corpses, cats, and elk, as well as tire, stones, and boxes.Keywords indicating the state of drainage such as clogging, stagnation, and flooding were mainly used for drainage facilities, and parking, banners, and garbage were analyzed to be frequently used for others.In the matching analysis, 79% were found to match accurately, and 21% were found to require reclassification.The results presented in this study can be used as basic data for the establishment of efficient road management policies by the Road Management Agency, and in particular, matching analysis results are expected to be able to accurately assign reporting types through learning technologies such as text mining and deep learning in the future for efficient