Categorization Architecture with Predictive Reasoning and Alignment for UNSPSC

Erli Wang, Dakshi Kapugama Geeganage, Opec Kemp

Abstract: The United Nations Standard Products and Services Code (UNSPSC) is a four-tier hierarchical taxonomy comprising segments, families, classes, and commodities, designed to facilitate globally consistent categorization of goods and services. The Queensland Government processes over seven million new procurement transactions each quarter. In line with recommendations from the Queensland Audit Office, all transactions are mapped to the UNSPSC hierarchy to support timely decision-making and improve operational efficiency. In this paper, we introduce CAPRA — the Categorization Architecture with Predictive Reasoning and Alignment — a framework for rigorously evaluating real-world labeling pipelines. CAPRA systematically compares multiple candidate configurations across four key dimensions: accuracy, model size, inference speed, and operational cost. To maintain data quality without overburdening human experts—who cannot feasibly validate millions of transactions each quarter—we embed an AI-driven feedback mechanism and a retrieval-based review process that selectively flag only the lowest-confidence predictions for expert adjudication. This “smart sampling” approach closes the data-service loop by focusing domain expertise where it matters most. Finally, we demonstrate CAPRA’s effectiveness through extensive experiments on Queensland Government expenditure data. The results show that CAPRA outperforms a fine-tuned Large Language Model (LLM) by 19% in prediction accuracy while requiring only 10% of its inference time, enabling seamless deployment in a production environment.