The use of machine learning to identify the correctness of HS Code for the customs import declarations

Abstract

As an increasing volume of international trade activities around the world, the amount of cross-boarder import declarations grows rapidly, resulting in an unprecedented scale of potentially fraudulent transactions, in particular false commodity code (e.g., HS Code). The incorrect HS Code will cause duty risk and adversely impact the revenue collection. Physical investigation by the customs administrations is impractical due to the substantial quantity of declarations. This paper provides an automatic approach by harnessing the power of machine learning techniques to relief the burden of customs targeting officers. We introduced a novel model based on the off-theshelf embedding encoder to identify the correctness of HS Code without any human effort. Determining whether the HS Code is correctly matched with commodity description is a classification task, so the labelled data is typically required. However, the lack of gold standard labelled data sets in customs domain limits the development of supervised-based approach. Our model is developed by the unsupervised mechanism and trained on the unlabelled historical declaration records, which is robust and able to be smoothly adapted by the different customs administrations. Rather than typically classifying whether the HS Code is correct or not, our model predicts the score to indicate the degree of the HS Code being correct. We have evaluated our proposed model on the ground-truth data set provided by Dutch customs officers. Results show promising performance of 71% overall accuracy.

Automatic Product Classification in International Trade: Machine Learning and Large Language Models

“…Ruder (2020) uses a variety of ML and deep learning models to classify product descriptions from the US Bill of Lading and reaches accuracy levels of approximately 60%. Chen et al (2021) apply unsupervised ML and an off-the-shelf embedding encoder to automatically assess whether reported HS codes in cross-border import declarations are correct. They achieve an overall success rate of 71% on an HS 6-digit dataset provided by Dutch customs Turhan et al (2015) adopt a different strategy whereby they use visual properties along with product labels and descrip- 6 We also tested the performance of GPT 3.5 in mapping sector descriptions onto the North American Industry Classification System (NAICS).…”

Modelling Customs Revenue in Ghana Using Novel Time Series Methods

“…Chen et al [11] in their work proposed an automatic approach by harnessing the power of machine learning techniques to relieve the burden of customs targeting officers who set revenue targets.…”