Research on the design of a data mining-based financial audit model for financial multi-type data processing and audit trail discovery
Lei Yu, Jifeng Wang · Systems and Soft Computing · 2025
There are many difficult to discover problems in the massive data of the financial field. Therefore, the purpose of the research is to extract valuable information from the massive financial data in depth and quickly, help auditors identify abnormal patterns and potential risks, and improve the intelligence level of auditing, reducing the impact of human errors and subjective judgments. The study validates the effectiveness of self-organising mapping neural network for voucher summarisation algorithm and bidirectional long and short-term memory based neural network; and designs a financial audit system that can process financial multiple data types while automatically discovering audit clues through data mining techniques based on financial audit trail discovery. The experiments conducted show that the clustering quality and efficiency of the self-organising mapping neural network for voucher summarisation algorithm has been improved, and the comparison experiments show that the improved two-way long and short-term memory neural network algorithm is effective with an accuracy rate of 95.4%, which is better than the accuracy rate of 90.35% of the traditional two-way long and short-term memory neural network algorithm. The convenient and effective financial audit system model visualises operational information during operation and can identify the nodes of operational errors, bringing convenience to the audit work.