Effective item categorization is vital for businesses, enabling the transformation of unstructured datasets into organized categories that streamline inventory management. Despite its importance, item categorization remains highly subjective and lacks a uniform standard across industries and businesses. The United Nations Standard Products and Services Code (UNSPSC) provides a standardized system for cataloguing inventory, yet employing UNSPSC categorizations often demands significant manual effort. This paper investigates the deployment of Large Language Models (LLMs) to automate the classification of inventory data into UNSPSC codes based on Item Descriptions. We evaluate the accuracy and efficiency of LLMs in categorizing diverse datasets, exploring their language processing capabilities and their potential as a tool for standardizing inventory classification. Our findings reveal that LLMs can substantially diminish the manual labor involved in item categorization while maintaining high accuracy, offering a scalable solution for businesses striving to enhance their inventory management practices.
Glitch Tokens in Large Language Models: Categorization Taxonomy and Effective Detection
Huazhong University of Science and Technology（华中科技大学）；Nanyang Technological University（南洋理工大学）；Virginia Tech（弗吉尼亚理工大学）；The University of New South Wales（新南威尔士大学）
Charting the European LLM Benchmarking Landscape: A New Taxonomy and a Set of Best Practices
Jožef Stefan Institute（乔塞夫·斯塔芬研究所）；University of Ljubljana（卢布尔雅那大学）；Faculty of Arts, University of Ljubljana（艺术学院，卢布尔雅那大学）；Faculty of Computer and Information Science, University of Ljubljana（计算机与信息科学学院，卢布尔雅那大学）；Institute of Contemporary History（现代史研究所）
Achieving Peak Performance for Large Language Models: A Systematic Review
Óbuda University（奥布达大学）；J. Selye University（塞耶大学）；Institute for Computer Science and Control (SZTAKI)（计算机科学与控制研究所（SZTAKI））；Hungarian Research Network (HUN-REN)（匈牙利研究网络（HUN-REN））
