Enhancing Data Integrity in Computerized Accounting Information Systems Using Supervised and Unsupervised Machine Learning Algorithms Implement A SEM-PLS Analysis

Abstract

The paper determines Machine Learning (ML) applications of both supervised and unsupervised types in computerised accounting information systems (CAIS) to improve data consistency. The Partial Least Squares Structural Equation Modelling (SEM-PLS) approach was used for the processing of data, which comprised 163 building companies in China that were using Building Information Modelling (BIM). This paper examines the financial data in view of ML algorithms and looks into the way ML improves financial data accuracy, consistency, reliability, and consistency. The results revealed that the integration of ML algorithms could increase data integrity (DI) by as much as 27% and detection of error by as much as 35% compared with manual methods. These results emphasise artificial intelligence (AI) solutions' leading role in improving CAIS-focused financial decision-making and operational control systems, demonstrating how AI can be applied to the field efficiently. By investigating the impact of ML on data integrity, which is measured through advanced SEM-PLS technique, this research is among the first studies on AI finance to deepen the knowledge.