Monitor and Detect Suspicious Transactions With Database Forensic Analysis

Abstract

The extensive usage of web has given rise to financially motivated illegal covert online transactions. So the digital investigators have approached databases for investigating undetected illegal transactions. The authors here have designed and developed a methodology to find the illegal financial transactions through the database logs. The objective is to monitor database transactions for detecting and reporting risk level of suspicious transactions. Initially, the process extracts SQL transactions from logs of different database systems, then transforms and loads them separately in uniform XML format which gives the transaction records and its metadata. The transaction records are processed with well-defined rules to get outliers present as suspicious transactions. This gives the initial belief of the transactions to be suspicious. The belief value of transactions is further rationalised using Dempster-Shafer's theory. This verifies the uncertainty and risk level of the suspected transactions to assure occurrences of fraud transactions.