Please use this identifier to cite or link to this item: https://www.um.edu.mt/library/oar/handle/123456789/23089
Title: A domain specific property language for fraud detection to support agile specification development
Authors: Calafato, Aaron
Colombo, Christian
Pace, Gordon J.
Keywords: Domain-specific programming languages
Natural language processing (Computer science)
Fraud investigation -- Data processing
Forensic accounting
Issue Date: 2014-11
Publisher: University of Malta. Faculty of ICT
Citation: Calafato, A., Colombo, C., & Pace, G. J. (2014). A domain specific property language for fraud detection to support agile specification development. Computer Science Annual Workshop CSAW’14, Msida. 1-2.
Abstract: Fraud detection is vital in any financial transaction system, including the collection of tax. The identification of fraud cases was traditionally carried out manually, having fraud experts going through their records and intuitively selecting the ones to be audited — a lengthy and unstructured process. Although work has been done with regards to the use of artificial technology for fraud pattern discovery, the results are not encouraging without major intervention by fraud experts [4]. Nowadays, in practice, patterns identified by fraud experts are coded by the software developers who select fraud cases from a database. The resulting application is verified by the fraud expert, who may feel the need to refine the rules in multiple iterations. However, this process is prone to human-induced bugs due to the continuous manual work. A better approach would include the description of rules through the use of a structured grammar, understandable by a computer system. With a compilable set of descriptions, the rules may be automatically processed against historical data — limiting the dependency on a software developer solely to the process of setting up the system.
URI: https://www.um.edu.mt/library/oar//handle/123456789/23089
Appears in Collections:Scholarly Works - FacICTCS

Files in This Item:
File Description SizeFormat 
Proceedings of CSAW’14 - A3.pdf95.01 kBAdobe PDFView/Open


Items in OAR@UM are protected by copyright, with all rights reserved, unless otherwise indicated.