A framework for detecting financial statement fraud through multiple data sources

Abstract

This project deals with how to detect fraud and non-compliance in financial statements in the present day in one of the biggest economies in the world, the U.S. Since it is mainly public companies that release detailed financial infor-mation, they are the focus. This project focuses on the top five market sectors where fraud is most common. It focuses on a variety of fraud types, but not on cases of deception that do not constitute fraud. A framework will be proposed which ac-counts for both structured data (the numbers in the balance sheet, income statement and cash flow statement) and unstruc-tured data (the footnotes in these financial statements). It uses ontology-driven data mining techniques to do so.