Policy improvement and research on maritime safety prevention and damage minimization are underway due to the annually increasing number of maritime accidents that are growing in scale. Generally, statistical data or casebooks from the Korea coast guard and the Korean Maritime Safety Tribunal are utilized for policy, business, or research processes involving accident prevention. These uses demonstrate the limitations of the data utilizing process for accident prevention. There are constraints in current statistical data articles for insightful analysis of accident dangerousness, and it is difficult to discover new trends through the correlation of each article. This study proposes a ‘maritime safety information classification system’, which enables a wider variety of data to be utilized for marine accident prevention related research, policy making, and decision making. The presented research will analyze the actual conditions of boating accidents from the perspective of initial reaction by applying the classification system. This cannot be found from existing statistical data. This is implemented with boating accident reports in the central sea from 2015-2018. The analysis includes the initial reaction related parts such as the accident report, propagation of the situation, dispatch (arrival time, etc.), report ratio and development by types accident types, and human life damage correlation due to it. It will then suggest an analyzed date utilization plan by government offices.