Demonstration of Public Inquiry Response Utilizing Public Inquiry Data, Customs

Overview

As overseas direct purchases and personal trade transactions increase, the need for individual consumers to understand customs administration and legal information is growing. Existing customer service methods have primarily relied on phone or in-person consultations, which struggle to effectively handle the rising demand for assistance. The project aims to develop an automated customer service system utilizing customs administration information and historical customer service data to reduce the workload on staff and improve service quality.

Project description

This project builds an AI-based automated response system using public inquiry data and customs administration information. During data collection, internal FAQ documents (8 documents) in JSON and PDF formats and webpage materials were utilized.

This data was preprocessed to generate SmartChat-based Q&A responses. After registering 3,500 query data points, the data underwent training and indexing processes using SearchGPT and SmartChat.

Subsequently, a demo page was developed for user testing to validate the effectiveness of the public inquiry response system.

Key technological features

• Service Development Utilizing Customs Administration Information and Historical Public Complaint Response Data (SearchGPT + SmartChat)
• Building an environment where users can easily search by consolidating scattered information (SearchGPT)
• Reducing the burden on civil petition handlers and improving the quality of customs administration services

Expected outcomes

Through this project, the Korea Customs Service and related agencies are expected to be able to respond to public inquiries more promptly and accurately. In particular, the AI chatbot based on SmartChat demonstrates response performance comparable to human-generated responses even for automatically generated queries, having learned from existing public inquiry data. The performance of such systems will enable customer service representatives to retrieve information more quickly and reduce their workload, while individual consumers can easily obtain desired customs administration and legal information through the search function. Ultimately, this project will enhance the quality of customer service and contribute to providing efficient administrative services.

Launch year: 2023

Attribution

Case study courtesy of the National Information Society Agency, Republic of Korea, Apolitical's Content Partner. Sourced from Use Cases of Public AI Service, Vol.1,  https://eng.nia.or.kr/site/nia_eng/ex/bbs/View.do?cbIdx=31975&bcIdx=28997&parentSeq=29027