Automated methods can streamline the analysis of audit reports
CER's PhD student Dennis Hedback investigates whether machine learning can be used to automatically identify audit reports where the auditor warns that a company is at risk of not being able to continue its operations.
The article "Identifying going concern audit opinions using supervised machine learning", published in the journal Intelligent Systems in Accounting, Finance and Management, highlights that such alerts are important, but can be very time-consuming and expensive to find and compile manually, especially when dealing with large numbers of companies. This can be a particular problem in research, where one often needs to analyze extensive data material as part of collecting data for other types of audit studies.
The results of the study show that machine learning models can determine whether an audit report contains a continued operation warning with high accuracy. The author therefore concludes that machine learning can make it relatively easy and inexpensive to identify continuation warnings in audit reports, especially for purposes where large numbers of audit reports need to be analyzed. This can be done with relatively simple models that can run quickly even on fairly weak hardware, without having to use more resource-intensive techniques such as large language models of the ChatGPT type.
Contact
