Multilabel Classification of Account Code in Double-Entry Bookkeeping
Pakorn Kotepuchai, Yachai Limpiyakorn · 2024
AI technology has fostered enterprises to stay strong contenders in today competitive world, whereas bookkeeping plays an indispensable role in financial management of successful business. Bookkeeping deals with recording financial transactions associated with account codes. Manually fill-in account codes would be error-prone and resource consumption, while entering the accurate account codes is essential for better decision-making and faster business growth. Automated suggestion of account codes with AI can save time, boost productivity and output quality by reducing human error. This paper thus presents an approach of AI-enabled system to suggest account codes in double-entry bookkeeping method. Natural language processing is mainly applied during data preprocessing for enabling machines to process and extract meaning from textual data. The models are trained with three machine learning algorithms including CART, Random Forest, and Multilayer Perceptron. The experimental results reported that Random forest model insignificantly outperformed the others.