ICCA-RAG: Intelligent Customs Clearance Assistant Using Retrieval-Augmented Generation (RAG)

Document processing and query generation tasks in customs declaration scenarios face key challenges such as the complexity of multimodal data, adaptability to dynamic regulations, and ambiguity in query semantics. This study proposes a Retrieval-Augmented Generation system (ICCA-RAG) that addresses...

Full description

Saved in:
Bibliographic Details
Main Authors: Rong Hu, Sen Liu, Panpan Qi, Jingyi Liu, Fengyuan Li
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10897992/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849772883976388608
author Rong Hu
Sen Liu
Panpan Qi
Jingyi Liu
Fengyuan Li
author_facet Rong Hu
Sen Liu
Panpan Qi
Jingyi Liu
Fengyuan Li
author_sort Rong Hu
collection DOAJ
description Document processing and query generation tasks in customs declaration scenarios face key challenges such as the complexity of multimodal data, adaptability to dynamic regulations, and ambiguity in query semantics. This study proposes a Retrieval-Augmented Generation system (ICCA-RAG) that addresses the core issues of processing complex customs documents and dynamically generating queries through multimodal document parsing, sparse-dense hybrid storage, and context-driven large language model generation. In terms of multimodal document parsing, the system supports comprehensive parsing of PDFs, images, tables, and text, which are uniformly transformed into semantic vectors and keyword indices for hybrid storage. By combining the retrieval and generation modules, the ICCA-RAG system achieves significant improvements in contextual relevance and generation accuracy. Compared to traditional methods, the ICCA-RAG system demonstrates a 20.1% increase in answer correctness, a 15.3% increase in answer relevancy, and an 18.7% increase in the faithfulness of generated content, with outstanding performance in noisy query scenarios. The research findings validate the ICCA-RAG system’s advancement and applicability in handling complex document processing and professional domain question-answering tasks, while also providing a transferable technical framework for other fields, such as law and healthcare.
format Article
id doaj-art-ffd4dd15d0d14dd8a42c8cf485e31eba
institution DOAJ
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-ffd4dd15d0d14dd8a42c8cf485e31eba2025-08-20T03:02:11ZengIEEEIEEE Access2169-35362025-01-0113397113972610.1109/ACCESS.2025.354440810897992ICCA-RAG: Intelligent Customs Clearance Assistant Using Retrieval-Augmented Generation (RAG)Rong Hu0https://orcid.org/0009-0002-9717-042XSen Liu1https://orcid.org/0009-0009-7272-8539Panpan Qi2Jingyi Liu3Fengyuan Li4Customs and Public Management College, Shanghai Customs University, Shanghai, ChinaDepartment of Electronic Information, Shanghai Dianji University, Shanghai, ChinaInformation Department, Xinglin College, Nantong University, Nantong, ChinaSchool of Information Engineering, Xi’an Jiaotong University, Xi’an, ChinaXi’an Jiaotong University, Xi’an, ChinaDocument processing and query generation tasks in customs declaration scenarios face key challenges such as the complexity of multimodal data, adaptability to dynamic regulations, and ambiguity in query semantics. This study proposes a Retrieval-Augmented Generation system (ICCA-RAG) that addresses the core issues of processing complex customs documents and dynamically generating queries through multimodal document parsing, sparse-dense hybrid storage, and context-driven large language model generation. In terms of multimodal document parsing, the system supports comprehensive parsing of PDFs, images, tables, and text, which are uniformly transformed into semantic vectors and keyword indices for hybrid storage. By combining the retrieval and generation modules, the ICCA-RAG system achieves significant improvements in contextual relevance and generation accuracy. Compared to traditional methods, the ICCA-RAG system demonstrates a 20.1% increase in answer correctness, a 15.3% increase in answer relevancy, and an 18.7% increase in the faithfulness of generated content, with outstanding performance in noisy query scenarios. The research findings validate the ICCA-RAG system’s advancement and applicability in handling complex document processing and professional domain question-answering tasks, while also providing a transferable technical framework for other fields, such as law and healthcare.https://ieeexplore.ieee.org/document/10897992/Customs declaration assistancedynamic regulation adaptationintelligent question-answering systemlarge language model (LLM)multimodal document parsingretrieval-augmented generation (RAG)
spellingShingle Rong Hu
Sen Liu
Panpan Qi
Jingyi Liu
Fengyuan Li
ICCA-RAG: Intelligent Customs Clearance Assistant Using Retrieval-Augmented Generation (RAG)
IEEE Access
Customs declaration assistance
dynamic regulation adaptation
intelligent question-answering system
large language model (LLM)
multimodal document parsing
retrieval-augmented generation (RAG)
title ICCA-RAG: Intelligent Customs Clearance Assistant Using Retrieval-Augmented Generation (RAG)
title_full ICCA-RAG: Intelligent Customs Clearance Assistant Using Retrieval-Augmented Generation (RAG)
title_fullStr ICCA-RAG: Intelligent Customs Clearance Assistant Using Retrieval-Augmented Generation (RAG)
title_full_unstemmed ICCA-RAG: Intelligent Customs Clearance Assistant Using Retrieval-Augmented Generation (RAG)
title_short ICCA-RAG: Intelligent Customs Clearance Assistant Using Retrieval-Augmented Generation (RAG)
title_sort icca rag intelligent customs clearance assistant using retrieval augmented generation rag
topic Customs declaration assistance
dynamic regulation adaptation
intelligent question-answering system
large language model (LLM)
multimodal document parsing
retrieval-augmented generation (RAG)
url https://ieeexplore.ieee.org/document/10897992/
work_keys_str_mv AT ronghu iccaragintelligentcustomsclearanceassistantusingretrievalaugmentedgenerationrag
AT senliu iccaragintelligentcustomsclearanceassistantusingretrievalaugmentedgenerationrag
AT panpanqi iccaragintelligentcustomsclearanceassistantusingretrievalaugmentedgenerationrag
AT jingyiliu iccaragintelligentcustomsclearanceassistantusingretrievalaugmentedgenerationrag
AT fengyuanli iccaragintelligentcustomsclearanceassistantusingretrievalaugmentedgenerationrag