An open data cleaning framework based on semantic rules for Continuous Auditing

Continuous Auditing (CA) is an important form of computer-assisted audit techniques (CAATs), which is also an active research domain in audit field. Because of the strict requirements for data quality for Continuous Auditing, a semantic rule-based open data cleaning framework (ODCF) with self-learning function is designed in this paper, which can improve the accuracy and adaptability of data cleaning. The semantic rules were used in the framework to recognize the hierarchy semantic and dependence among the fields. Firstly, introduce the structure, components and workflow of the framework in detail. And then describe the cooperation of various components, which improves cleaning efficiency, through the processing of various types of dirty data. Finally, analyze the performance of the framework, and point out its adaptability and universality to use. This is an open framework for data cleaning with good scalability, which will becomes more and more perfect and powerful with the success of data cleaning practice in different fields.