No abstract

No abstract

Corporate fraud these days represents a huge cost to our economy.Academic literature already concentrated on how data mining techniques can be of value in the fight against fraud. All this research focusses on fraud detection, mostly in a context of external fraud. In this paper we discuss the use of a data mining approach to reduce the risk of internal fraud. Reducing fraud risk comprehends both detection and prevention, and therefore we apply descriptive data mining as opposed to the widely used prediction data mining techniques in the literature. The results of using a multivariate latent class clustering algorithm to a case company's procurement data suggest that applying this technique in a descriptive data mining approach is useful in assessing the current risk of internal fraud. The same results could not be obtained by applying a univariate analysis.

There is a large body of accounting research literature examining the use of analytical procedures by auditors and proposing either new types of analytical procedures or more effective ways of implementing existing procedures. In this paper, we demonstrate-using procurement data from a leading global bank-the value added in an audit setting of a new type of analytical procedure: process mining of event logs. In particular, using process mining, we are able to identify numerous transactions that we consider to be audit-relevant information, including payments made without approval, violations of segregation of duty controls, and violations of company-specific internal procedures. Furthermore, these identified anomalies were not detected by the bank's internal auditors when they conducted their examination of that same data using conventional audit procedures, thus establishing the benefits of using process mining to complement existing audit methods. Process mining is a very different approach to evidence collection and analysis as it does not focus on the value of transactions and its aggregations, but on the transactional processes themselves. In addition to demonstrating the benefits of process mining in an audit context, this paper also discusses the contributions that process mining can make both to accounting research and auditing practice.

Over the last decades, the field of process mining has emerged as a response to a growing amount of event data being recorded in the context of business processes. Concurrently with the increasing amount of literature produced in this field, a set of tools has been developed to implement the various algorithms and provide them to end users. However, the majority of tools does not provide the possibility of creating workflows which can be reused at a later point in time to reproduce the results, and most tools are not easily customizable. This paper introduces bupaR, an integrated collection of Rpackages which creates a framework for reproducible process analysis in R and supports different steps of a process analysis project, from data extraction to data analysis. It is an extensible framework of several R-packages to analyse process data, each with their specific purpose and set of tools.