A Deterministic Agentic Workflow for HS Tariff Classification: Multi-Dimensional Rule Reasoning with Interpretable Decisions

2026-05-14

AI Summary

This study addresses the challenge of priority reasoning among multidimensional rules—such as material, form, and function—in mapping product descriptions to Harmonized System (HS) codes. The authors propose a deterministic agent workflow that employs a fixed control flow to invoke large language models in distinct stages, embedding structured knowledge from China’s HS tariff nomenclature to produce interpretable and institutionally coherent classification decisions. The approach integrates a six-stage online inference pipeline with offline knowledge engineering, leveraging constrained calls and localized validation using the Qwen3.6 model series. Evaluated on the HSCodeComp dataset, the method achieves 77.4% top-1 accuracy at the six-digit level (using Qwen3.6-27B-FP8) and 84.2% at the four-digit level. Manual auditing reveals that certain ground-truth annotations deviate from the General Rules for the Interpretation of the HS.

Abstract

Harmonized System (HS) tariff classification is a high-stakes, expert-level task in which a free-form product description must be mapped to a specific six- or eight-digit code under the General Interpretive Rules (GIR), section notes, chapter notes, and Explanatory Notes. The difficulty lies not in knowledge volume but in *multi-dimensional rule reasoning*: a correct classification must satisfy competing priority rules along several axes simultaneously, including material, form, function, essential character, the part-versus-whole boundary, and specific listing versus residual headings. End-to-end prompting of large language models fails characteristically by resolving one axis while ignoring the priority constraints on the others. We present a *deterministic agentic workflow* in contrast to self-planning agents: the control flow is fixed, language model calls are confined to narrow stages, and reflection and verification are retained as local mechanisms. This design yields interpretability by construction--each decision is decomposed into stage-wise structured outputs with verbatim citation of the chapter or section notes that bear on it. The architecture combines offline knowledge-engineering of the Chinese HS tariff with an online six-stage pipeline. Evaluated on HSCodeComp at the six-digit level, the workflow reaches 75.0% top-1 and 91.5% top-3 at four digits, and 64.2% top-1 and 78.3% top-3 at six digits with Qwen3.6-plus; an open-weight Qwen3.6-27B-FP8 backbone in non-thinking mode achieves 84.2% four-digit and 77.4% six-digit top-1 agreement with the frontier model. A two-stage manual audit of 226 six-digit disagreements suggests that a non-trivial fraction of HSCodeComp ground-truth labels may deviate from HS general rules; full adjudication records are released in the appendix as preliminary findings for community review.

Problem

• Harmonized System
• tariff classification
• multi-dimensional rule reasoning
• priority constraints
• interpretability

Innovation

• deterministic agentic workflow
• multi-dimensional rule reasoning
• interpretable decisions
• Harmonized System classification
• structured reasoning pipeline