skip to main content
research-article

Explainable Product Classification for Customs

Published: 22 February 2024 Publication History

Abstract

The task of assigning internationally accepted commodity codes (aka HS codes) to traded goods is a critical function of customs offices. Like court decisions made by judges, this task follows the doctrine of precedent and can be nontrivial even for experienced officers. Together with the Korea Customs Service (KCS), we propose a first-ever explainable decision supporting model that suggests the most likely subheadings (i.e., the first six digits) of the HS code. The model also provides reasoning for its suggestion in the form of a document that is interpretable by customs officers. We evaluated the model using 5,000 cases that recently received a classification request. The results showed that the top-3 suggestions made by our model had an accuracy of 93.9% when classifying 925 challenging subheadings. A user study with 32 customs experts further confirmed that our algorithmic suggestions accompanied by explainable reasonings, can substantially reduce the time and effort taken by customs officers for classification reviews.

References

[1]
Fatma Altaheri and Khaled Shaalan. 2020. Exploring machine learning models to predict harmonized system code. In Proc. of the European, Mediterranean, and Middle Eastern Conference on Information Systems. 291–303.
[2]
Boli Chen, Xin Huang, Lin Xiao, Zixin Cai, and Liping Jing. 2020. Hyperbolic interaction model for hierarchical multi-label classification. In Proc. of the AAAI Conference on Artificial Intelligence. 7496–7503.
[3]
Xi Chen, Stefano Bromuri, and Marko van Eekelen. 2021. Neural machine translation for harmonized system codes prediction. In Proc. of the International Conference on Machine Learning Technologies (ICMLT). 158–163.
[4]
Hao Cheng, Xiaoqing Yang, Zang Li, Yanghua Xiao, and Yucheng Lin. 2019. Interpretable text classification using CNN and max-pooling. arXiv:1910.11236. Retrieved from https://arxiv.org/abs/cs/1910.11236
[5]
Huyen Chip. 2022. Designing Machine Learning Systems: An Iterative Process for Production-Ready Applications. (1 ed.), O’Reilly.
[6]
Ciel HS. 2022. Harmonized commodity description and coding system explanatory notes. Retrieved from https://www.clhs.co.kr/uploads/lawfile/404n2563.pdf. (2022). Accessed: 2022-02-27.
[7]
Kevin Clark, Minh-Thang Luong, Quoc V Le, and Christopher D Manning. 2020. ELECTRA: Pre-training text encoders as discriminators rather than generators. In Proceedings of the International Conference on Learning Representations (ICLR).
[8]
Tirthankar Dasgupta, Rupsa Saha, Lipika Dey, and Abir Naskar. 2018. Automatic extraction of causal relations from text using linguistically informed deep neural networks. In Proceedings of the Annual Meeting of the Special Interest Group on Discourse and Dialogue (SIGDIAL). 306–316.
[9]
Hans de Bruijn, Martijn Warnier, and Marijn Janssen. 2021. The perils and pitfalls of explainable AI: Strategies for explaining algorithmic decision-making. Government Information Quarterly 39, 2 (2021), 101666.
[10]
Liya Ding, ZhenZhen Fan, and DongLiang Chen. 2015. Auto-categorization of HS code using background net approach. Procedia Computer Science 60, 1 (2015), 1462–1471. DOI:DOI:
[11]
Yuwei Fang, Siqi Sun, Zhe Gan, Rohit Pillai, Shuohang Wang, and Jingjing Liu. 2020. Hierarchical graph network for multi-hop question answering. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP). 8823–8838.
[12]
Dirk Groeneveld, Tushar Khot, Mausam, and Ashish Sabharwal. 2020. A simple yet strong pipeline for HotpotQA. In Proc. of the Conference on Empirical Methods in Natural Language Processing (EMNLP). 8839–8845.
[13]
Chuan Guo, Geoff Pleiss, Yu Sun, and Kilian Q. Weinberger. 2017. On calibration of modern neural networks. In Proceedings of the International Conference on Machine Learning (ICML). 1321–1330.
[14]
Jacob Devlin Ming-Wei Chang Kenton and Lee Kristina Toutanova. 2019. BERT: Pre-training of deep bidirectional transformers for language understanding. In Proceedings of the Annual Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT). 4171–4186.
[15]
Tushar Khot, Peter Clark, Michal Guerquin, Peter A. Jansen, and Ashish Sabharwal. 2020. QASC: A dataset for question answering via sentence composition. In Proceedings of the AAAI Conference on Artificial Intelligence. 8082–8090.
[16]
Sundong Kim, Tung-Duong Mai, Sungwon Han, Sungwon Park, D. K. Thi Nguyen, Jaechan So, Karandeep Singh, and Meeyoung Cha. 2023. Active learning for human-in-the-loop customs inspection. IEEE Transactions on Knowledge and Data Engineering 35, 12 (2023), 12039–12052.
[17]
Sun Kim, Nicolas Fiorini, W. John Wilbur, and Zhiyong Lu. 2017. Bridging the gap: Incorporating a semantic similarity measure for effectively mapping PubMed queries to documents. Journal of Biomedical Informatics 75 (2017), 122–127.
[18]
Markus Langer, Daniel Oster, Timo Speith, Holger Hermanns, Lena Kästner, Eva Schmidt, Andreas Sesing, and Kevin Baum. 2021. What do we want from explainable artificial intelligence (XAI)?–A stakeholder perspective on XAI and a conceptual model guiding interdisciplinary XAI research. Artificial Intelligence 296, 3 (2021), 103473.
[19]
Dongju Lee, Gunwoo Kim, and Keunho Choi. 2020. CNN-based recommendation model for classifying HS code. Management & Information Systems Review 39, 3 (2020), 1–16.
[20]
Eunji Lee, Sundong Kim, Sihyun Kim, Sungwon Park, Meeyoung Cha, Soyeon Jung, Suyoung Yang, Yeonsoo Choi, Sungdae Ji, Minsoo Song, and Heeja Kim. 2021. Classification of goods using text descriptions with sentences retrieval. In Proceedings of the Korea Artificial Intelligence Conference (KAIA).
[21]
Guo Li and Na Li. 2019. Customs classification for cross-border e-commerce based on text-image adaptive convolutional neural network. Electronic Commerce Research 19, 4 (2019), 779–800.
[22]
Jeffrey Luppes, Arjen P de Vries, and Faegheh Hasibi. 2019. Classifying short text for the harmonized system with convolutional neural networks. Radboud University (2019).
[23]
Tung-Duong Mai, Kien Hoang, Aitolkyn Baigutanova, Gaukhartas Alina, and Sundong Kim. 2021. Customs fraud detection in the presence of concept drift. In Proceedings of the ICDM IncrLearn Workshop.
[24]
Carolyn McKay. 2020. Predicting risk in criminal procedure: Actuarial tools, algorithms, AI and judicial decision-making. Current Issues in Criminal Justice 32, 1 (2020), 22–39.
[25]
Minkyu Park. 2019. A study on the customs classification fallacy of certain ITA goods. Korea Trade Review 44, 2 (2019), 189–202.
[26]
Sungjoon Park, Jihyung Moon, Sungdong Kim, Won Ik Cho, Jiyoon Han, Jangwon Park, Chisung Song, Junseong Kim, Yongsook Song, Taehwan Oh, Joohong Lee, Juhyun Oh, Sungwon Lyu, Younghoon Jeong, Inkwon Lee, Sangwoo Seo, Dongjun Lee, Hyunwoo Kim, Myeonghwa Lee, Seongbo Jang, Seungwon Do, Sunkyoung Kim, Kyungtae Lim, Jongwon Lee, Kyumin Park, Jamin Shin, Seonghyun Kim, Lucy Park, Alice Oh, Jungwoo Ha, and Kyunghyun Cho. 2021. KLUE: Korean language understanding evaluation. In Proceedings of the Neural Information Processing Systems (NeurIPS), Track on Datasets and Benchmarks. 1–25.
[27]
Juan Ramos. 2003. Using TF-IDF to determine word relevance in document queries. In Proceedings of the International Conference on Machine Learning (ICML), Vol. 242. 29–48.
[28]
Minjoon Seo, Aniruddha Kembhavi, Ali Farhadi, and Hannaneh Hajishirzi. 2017. Bidirectional attention flow for machine comprehension. In Proceedings of the International Conference on Learning Representations (ICLR).
[29]
Korea Customs Service. 2018. E-commerce goods import trend. Retrieved from https://url.kr/4bp9ew. (2018). Accessed: 2023-07-16.
[30]
Jiaming Shen, Wenda Qiu, Yu Meng, Jingbo Shang, Xiang Ren, and Jiawei Han. 2021. TaxoClass: Hierarchical multi-label text classification using only class names. In Proceedings of the Annual Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT). 4239–4249.
[31]
Tania Sourdin. 2018. Judge v robot?: Artificial intelligence and judicial decision-making. University of New South Wales Law Journal 41, 4 (2018), 1114–1133.
[32]
Mokanarangan Thayaparan, Marco Valentino, Viktor Schlegel, and André Freitas. 2019. Identifying supporting facts for multi-hop question answering with document graph networks. In Proceedings of the Workshop on Graph-Based Methods for Natural Language Processing (TextGraphs). 42–51.
[33]
The Korea Times. 2015. Smartwatch is a communication device. Retrieved from https://tinyurl.com/4vrfx7ef. (2015). Accessed: 2022-02-27.
[34]
Santosh T.Y.S.S, Shanshan Xu, Oana Ichim, and Matthias Grabmair. 2022. Deconfounding legal judgment prediction for european court of human rights cases towards better alignment with experts. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP). 1120–1138.
[35]
Marco Valentino, Mokanarangan Thayaparan, and André Freitas. 2021. Unification-based reconstruction of multi-hop explanations for science questions. In Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics (EACL).
[36]
Hai Wang, Dian Yu, Kai Sun, Jianshu Chen, Dong Yu, David McAllester, and Dan Roth. 2019. Evidence sentence extraction for machine reading comprehension. In Proceedings of the Conference on Computational Natural Language Learning (CoNLL). 696–707.
[37]
Wenhui Wang, Nan Yang, Furu Wei, Baobao Chang, and Ming Zhou. 2017. Gated self-matching networks for reading comprehension and question answering. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL). 189–198.
[38]
Zhengyang Wang, Xia Hu, and Shuiwang Ji. 2020. iCapsNets: Towards interpretable capsule networks for text classification. arXiv:2006.00075. Retrieved from https://arxiv.org/abs/cs/2006.00075
[39]
Sarah Wiegreffe and Ana Marasović. 2021. Teach me to explain: A review of datasets for explainable NLP. In Proceedings of the Neural Information Processing Systems (NeurIPS), Track on Datasets and Benchmarks.
[40]
Wikipedia. 2022. General rules for the interpretation of the harmonized system. Retrieved from https://en.wikipedia.org/wiki/General_Rules_for_the_Interpretation_of_the_Harmonized_System. (2022). Accessed: 2022-02-27.
[41]
Christoph Winter. forthcoming. The challenges of artificial judicial decision making for liberal democracy. In Proceedings of the Judicial Decision-Making: Integrating Empirical and Theoretical Perspectives.
[42]
World Customs Organization. 2018. HS compendium—The harmonized system, a universal language for international trade. Retrieved from https://tinyurl.com/ycr259ty. (2018). Accessed: 2022-02-27.
[43]
Vikas Yadav, Steven Bethard, and Mihai Surdeanu. 2020. Unsupervised alignment-based iterative evidence retrieval for multi-hop question answering. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL). 4514–4525.
[44]
Zhilin Yang, Peng Qi, Saizheng Zhang, Yoshua Bengio, William Cohen, Ruslan Salakhutdinov, and Christopher D. Manning. 2018. HotpotQA: A dataset for diverse, explainable multi-hop question answering. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP). 2369–2380.
[45]
Yu Zhang, Zhihong Shen, Yuxiao Dong, Kuansan Wang, and Jiawei Han. 2021. MATCH: Metadata-aware text classification in a large hierarchy. In Proceedings of the Web Conference (WWW). 3246–3257.
[46]
Wei Zhao, Rahul Singh, Tarun Joshi, Agus Sudjianto, and Vijayan N. Nair. 2021. Self-interpretable convolutional neural networks for text classification. arXiv:2105.08589. Retrieved from https://arxiv.org/abs/cs/2105.08589

Cited By

View all
  • (2025)Sequential gated recurrent and self attention explainable deep learning model for predicting hydrogen production: Implications and applicabilityApplied Energy10.1016/j.apenergy.2024.124851378(124851)Online publication date: Jan-2025
  • (2024)Pruned tree-structured temporal convolutional networks for quality variable prediction of industrial processJournal of Process Control10.1016/j.jprocont.2024.103312143(103312)Online publication date: Nov-2024
  • (2024)Deep learning-based time series forecastingArtificial Intelligence Review10.1007/s10462-024-10989-858:1Online publication date: 25-Nov-2024

Recommendations

Comments

Information & Contributors

Information

Published In

cover image ACM Transactions on Intelligent Systems and Technology
ACM Transactions on Intelligent Systems and Technology  Volume 15, Issue 2
April 2024
481 pages
EISSN:2157-6912
DOI:10.1145/3613561
  • Editor:
  • Huan Liu
Issue’s Table of Contents

Publisher

Association for Computing Machinery

New York, NY, United States

Publication History

Published: 22 February 2024
Online AM: 01 December 2023
Accepted: 17 November 2023
Revised: 02 October 2023
Received: 20 April 2022
Published in TIST Volume 15, Issue 2

Permissions

Request permissions for this article.

Check for updates

Author Tags

  1. Product classification
  2. interpretability
  3. decision support
  4. human-centered explainable AI

Qualifiers

  • Research-article

Funding Sources

  • Institute for Basic Science
  • NRF
  • IITP
  • Ministry of Science and ICT in Korea

Contributors

Other Metrics

Bibliometrics & Citations

Bibliometrics

Article Metrics

  • Downloads (Last 12 months)358
  • Downloads (Last 6 weeks)34
Reflects downloads up to 12 Feb 2025

Other Metrics

Citations

Cited By

View all
  • (2025)Sequential gated recurrent and self attention explainable deep learning model for predicting hydrogen production: Implications and applicabilityApplied Energy10.1016/j.apenergy.2024.124851378(124851)Online publication date: Jan-2025
  • (2024)Pruned tree-structured temporal convolutional networks for quality variable prediction of industrial processJournal of Process Control10.1016/j.jprocont.2024.103312143(103312)Online publication date: Nov-2024
  • (2024)Deep learning-based time series forecastingArtificial Intelligence Review10.1007/s10462-024-10989-858:1Online publication date: 25-Nov-2024

View Options

Login options

Full Access

View options

PDF

View or Download as a PDF file.

PDF

eReader

View online with eReader.

eReader

Full Text

View this article in Full Text.

Full Text

Figures

Tables

Media

Share

Share

Share this Publication link

Share on social media