Forensic Research & Technology

Biography

Biography: Zabihollah Rezaee

Abstract

We examine the adoption of the Big Data Framework in forensic accounting for detecting financial statement fraud (FSF) and the relevance and efficacy of Data Science principles and predictive algorithms in detecting, predicting, and preventing FSF. Business organizations lose about 5% of their revenues to fraud each year, which can exceed 3.5 trillion (USD) worldwide. The existence and persistence of financial statement fraud (FSF) is detrimental to the safety, soundness, and efficiency of our financial markets. Our study is intended to improve audit efficacy in discovering FSF by using Big Data and data analytics and algorithms. We conduct our analyses in four stages. In the first stage, we capture as much data as possible from a variety of sources to identify factors and potential signals that can lead to FSF. In the second stage we focus on identifying specific phenomena, characteristics, and symptoms that could trigger fraudulent incidents. In the third stage, we process the data using the Big Data Apache Spark platform to process the captured data and turn them into usable and relevant information in predicting and detecting FSF. In the fourth stage we use a Data Science approach and develop a suite of machine learning algorithms to estimate the probability of FSF occurrence. Big Data requires the use of sophisticated analytical tools and platforms such as Apache Spark to effectively and accurately identify potential risks that may trigger FSF. Corporations should proactively search for irregularities in Big Data and manage their risk profile in discovering FSF