An Intelligent Bookkeeping and Tax Filing Model for Farming Households Based on AI and Accounting Rules

With the advancement of digital village construction, farm households are increasingly operating online but commonly face practical difficulties such as disordered bookkeeping and cumbersome tax filing. Conventional enterprise-level fiscal and tax software, however, cannot accommodate the scenarios of farm households. To address this gap, this study integrates accounting principles, agriculture-related tax policies, and AI technologies such as optical character recognition (OCR) and natural language processing (NLP) to construct a lightweight, highly compliant prototype for intelligent bookkeeping and tax filing by farm households, and implements it on a WeChat mini-program platform. Retrospective simulation tests using historical operating data from 217 farm households show that the prototype significantly improves bookkeeping standardization rates and tax filing accuracy while effectively reducing fiscal and tax compliance risks. The standardized operating data it generates also provide a foundation for building a digital credit system for farm households. This study offers a reference for digital village construction and agriculture-related inclusive fiscal and tax services.