Using Data Mining Techniques for Fraud Detection in The Non-banking Sector

Abstract

A method known as data mining is used to extract knowledge and insights from vast amounts of data. To find different correlational patterns several computational and statistical techniques can be used. Data mining and other tools and techniques, such as artificial intelligence, can be used to do this activity. It tries to reduce the monetary losses these kinds of operations bring and ensure the business complies with all applicable legal and regulatory obligations. To stop fraud from happening, organizations must be able to recognize it before it occurs. This paper explores the various facets of fraud detection and how businesses may use it to stop it from happening. Following that, we discuss the many methods employed in this procedure, including clustering, unsupervised learning, and neural networks. In addition, we discuss the various data pretreatment methods employed in fraud detection. These include feature selection, data normalization, and extraction. Data visualization is crucial for deciphering and comprehending mining analysis outcomes. The study then discusses the numerous data mining applications for fraud detection. These include insurance, financial statement, credit card, and healthcare fraud. We give instances of when these methods have been used to spot fraudulent activity. The limitations of data mining for fraud detection are then discussed. The significance of this technology in the battle against financial crime is highlighted in this article, which offers a thorough review of the various facets of these sectors.

Financial Statement Text Information Mining and Key Information Extraction Model Construction

“…The researcher made significant contributions to the development of information retrieval systems capable of handling the complexity and amount of financial textual data. [16]. Data mining techniques have also been used to extract actionable insights from textual data in financial accounts.…”