Research on the Construction and Optimization of Enterprise Internal Audit Informatization System Based on Deep Learning Algorithm

Abstract

Traditional auditing methods are inefficient, making it easy for organizations to miss key risk points and difficult to uncover the potential risks behind the data. In this study, we construct and optimize an enterprise internal audit informatization system based on deep learning algorithms, which contains core modules such as data collection and preprocessing, deep learning model construction, audit decision-making and risk warning, etc. The modules work together to break the status quo of traditional audit data silos, further enhance the enterprise audit efficiency, and strengthen risk prevention and control. At the same time, the application of the system in large manufacturing enterprises A, prompting its audit efficiency by 60%, audit cycle significantly compressed, labor costs cut by 30%; audit quality significantly advanced, misjudgment, omission risk reduction of 40%, significantly reduce financial fraud, connected transactions and other hidden risks; cost control results, the potential risk of loss reduction of tens of millions of dollars per year. This research provides strong support for the transformation of the entire auditing ecology to digitalization and intelligence, injects new technological vitality into the knowledge system of the auditing discipline.