arXiv cs.CL· Souvik Nath, Sumit Wadhwa, Luis Perez·· 12 小时前AI 评分22
领域自适应小型语言模型用于结构化税码预测
Domain-Adaptive Small Language Models for Structured Tax Code Prediction
AI 导读
一篇 arXiv 论文提出基于编码器-解码器架构的领域自适应小型语言模型（SLM），用于从非结构化产品与服务数据中预测层级化税码序列。实验显示，该模型在 HSN 税码预测上优于扁平分类器，并在结构化序列生成任务上优于纯解码器和纯编码器架构。该方法可扩展至 UNSPSC、巴西 NCM 等其他政府强制税则编码。
正文 · 原文
Abstract:Every day, multinational firms process thousands of transactions, each of which must adhere to tax regulations that vary by jurisdiction and are often nuanced. The determination of product and service tax codes, such as HSN or SAC is a major use case in Tax compliance. An accurate determination of such codes is imperative to avoid any tax penalties. This paper proposes a domain-adaptive small language model (SLM) with an encoder-decoder architecture for the enhanced prediction of product and service tax codes. In this approach, we address the problem of predicting hierarchical tax code sequences using unstructured product and services data. We employ an SLM based upon encoder-decoder architecture as this enables sequential generation of tax codes to capture the hierarchical dependencies present within the tax codes. Our experiments demonstrate that encoder-decoder SLMs can be successfully applied to the sequential prediction of structured tax codes, a domain that remains comparatively unexplored in current NLP research. In this paper, we demonstrate the superior performance of the domain-adaptive encoder-decoder SLMs over flat classifiers when applied to the Harmonized System of Nomenclature (HSN), and achieve superior results compared to decoder-only and encoder-only architectures for structured sequence generation tasks. This approach can also be scaled to other government-mandated tax commodity codes, such as United Nations Standard Products and Services Codes (UNSPSC), or Brazil's Nomenclatura Comum do Mercosul (NCM).
| Comments: | 8 pages, 3 figures, 4 tables. v3: added related work, ablation study, and cost analysis. This work has been submitted to the IEEE for possible publication |
| Subjects: | Machine Learning (cs.LG); Computation and Language (cs.CL) |
| Cite as: | arXiv:2507.10880 [cs.LG] |
| (or arXiv:2507.10880v3 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2507.10880 arXiv-issued DOI via DataCite |
Submission history
From: Sumit Wadhwa [view email]
[v1]
Tue, 15 Jul 2025 00:46:01 UTC (541 KB)
[v2]
Sat, 19 Jul 2025 21:12:12 UTC (541 KB)
[v3]
Tue, 3 Nov 2026 21:11:26 UTC (1,063 KB)
来源：arXiv cs.CL · arxiv.org