인공지능 기반 해외직구 플랫폼의 관세 및 통관 효율성 제고 방안에 관한 연구

With the explosive global growth of he e-commerce market, the number of cross-border shopping transactions utilizing duty-free benefits is increasing every year. This increase in transaction volume, coupled with the import of low-priced products containing hazardous ingredients exceeding standards, is also raising safely concerns for Koreans. However, due to domestic customs conditions, most small-quantity items imported through cross-border shopping platforms are imported through simplified customs clearance methods, such as mail or list-based procedures. Consequently, it is difficult to deploy and manage sufficiently experienced professionals to quickly sort large volumes of general and hazardous goods. This leads to customs clearance delays and increased logistics costs, which can negatively impact profitability both domestically and internationally, making it difficult to provide faster and more accurate customs logistics services. While most academic discussions on customs clearance for cross-border shopping focus on innovative improvements to existing systems, research on hub-based customs clearance platforms, a new technology, in acceptable. Therefor, this paper prioritizes the application of AI technology to enhance the efficiency of customs clearance for the ever-increasing volume of cross-border shopping. This paper utilize data on customs clearance agencies related to cross-border shopping and utilizes Python, an AI-related programming language. To understand the current situation and build an AI-based e-commerce platform, we utilized data from cross-border shopping customs clearance agencies and implemented Python for coding logic. The research results suggest volume distribution and priority queue based customs clearance processing to address customs clearance delays caused by seasonal surges in volumes. Furthermore, to address the increased number of safety inspection items, we propose coding optimization for the selection of hazardous goods.