Nowadays, modern technology represented by unstructured text data information is widely used in all walks of life. According to the company’s unstructured text data information, the establishment of the most advanced company’s financial accounting artificial intelligence model has gradually become the yearning goal of Chinese companies to carry out financial work. In order to solve many problems such as low efficiency of traditional financial accounting and insufficient risk early warning ability, this paper used the Naive Bayes algorithm based on the unstructured text data of enterprises. It can improve the early warning accuracy of enterprise financial analysis and solve the problem of low efficiency of financial accounting, thereby promoting the transformation of financial accounting into an artificial intelligence model. By comparing the accuracy of financial risk early warning with the help of Naive Bayes algorithm and traditional manual mode, it was concluded that the Naive Bayes model algorithm based on enterprise unstructured text data had higher accuracy in corporate financial risk early warning, generally around 98%. It improved the accuracy by about 10% compared with the traditional manual mode, which was conducive to the continuous enhancement of financial accounting data collection, processing, and analysis capabilities.
