期刊名称：International Journal of Computer Trends and Technology
电子版ISSN：2231-2803
出版年度：2017
卷号：51
期号：1
页码：1-9
DOI：10.14445/22312803/IJCTT-V51P101
出版社：Seventh Sense Research Group
摘要：Fraud detection has been a major challenge in the financial industry. The fraud menace has made a lot of organization lost billions of naira especially in multidivisional and multibranch enterprises. Hence, there is a need for pragmatic approach to proffer solution to this challenge. The methodological approach used in this work involves the development of framework which includes the remote extraction of financial Journal Entries (JE) from each branch location of multidivisional multibranch enterprises and integrated into an SQL Server database using a standard data format through SQL Server Management Studio. The extracted data is thereafter used to build a central data warehouse that is transmitted to the auditors at the corporate headquarters of the enterprise using web applications tools through Internet. A Decision Tree data mining algorithm constructed is applied by the auditors at the corporate headquarters on the data warehouse to detect possible financial JE fraud. The tasks are guided by the concept of a three tier client/server architecture in which the data extraction and data warehouse construction tasks constitute a data/backend tier, the transmission of the data warehouse through the web services constitutes the application/ middle tier while the decision tree data mining algorithm application for fraud detection through a user interface program of Active Server Pages (ASP.NET) constitute a presentation tier. Therefore, supervised predictive machine learning was employed in this study because the classes for fraud detection are user defined, they are ensured to conform to the classification hierarchy of the Journal Entries (JE). The use of training data improves the ability to differentiate between classes with similar journal profiles using the methods that are more reliable and produce more accurate results.
关键词：Data Mining; Data Warehouse; Decision Tree; Journal Entries; Machine Learning; Web Services.
