초록 열기/닫기 버튼
Idle resources that have lost their original functions are classified into various areas such as space, knowledge, transportation, and objects based on sharing. In particular, idle spaces generated in farming and fishing villages are likely to be criminalized and abused in the long term, which is likely to cause urban aesthetics and social damage due to unauthorized dumping and various wastes. In addition, it is said that there is a high possibility of social problems due to the loss of economic function due to the decline in vitality of local commercial districts. Since idle spaces are likely to be used depending on the cause of occurrence, measures are needed to prevent them from being left unattended for a long time. In this study, unlike qualitative studies such as policy proposals, plans, strategies, and utilization methods conducted in previous studies, empirical analysis was conducted based on actual data. In this work, we intend to establish and propose an idle space prediction model to explore and efficiently manage available idle spaces. As a result of the experiment of this study, the optimal model was selected from the model using XGBoost among various machine learning techniques with an accuracy of 85.2%, and then artificial intelligence that can be explained was implemented through the SHAP technique. In this study, it has the following meanings. First, empirical analysis using public property data, second, idle space prediction model using data-based machine learning technique, third, idle space utilization through XAI implementation was attempted first, and fourth, it is meaningful in that efficient idle space management in Gyeongsangnam-do is possible. The model proposed in this study can be applied to the use of idle spaces in other regions, and is expected to be optimal operation and efficient management.
