Data mining approach to internal fraud in a project-based organization

Data mining is an efficient methodology for uncovering and extracting information from large databases, which widely used in different areas, e.g., customer relation management, financial fraud detection, healthcare manufacturing. has been successfully various detection prevention such as credit card fraud, taxation fund transfer fraud. However, there are insufficient researches about the usage of data related to internal control. In order increase awareness usefulness control, we developed a case study project-based organization. We analyze dataset working-hour claims projects, using two techniques: chi-square automatic interaction (CHAID) decision tree link analysis, describe characteristics fraudulent develop model potentially ones. Results indicate that following suspected working-hours claim were most significant: sector customer, origin level expertise consultant, cost consulting services. Our research contributes area control supported by mining, with goal prevent organizations.