Descriptive Data Mining on Fraudulent Online Dating Profiles

The increasing ease of access to the World Wide Web and email harvesting tools has enabled spammers to target a wider audience. The problem is where scams are widely encountered in day to day environment to individuals from all walks of life and result in millions of dollars in financial loss as well as emotional trauma (Newman 2005). This paper aims to analyse and examine the structure of Romance Fraud, in a bid to understand and detect Romance Fraud profiles. We focus on scams that utilise the medium of dating websites. The primary indicators of Romance Fraud identified in the literature include social factors, scam characteristics and content. The approach followed is informed by interpretivist and quantitative research perspectives. From this understanding, Romance Fraud can be viewed as a methodical attempt by scammers to penetrate their victims’ defences by impersonating someone else or creating an avatar (fictitious identity). A quantitative approach was undertaken in order to extract reflective, informative and rich data (Neuman 2003). The research methodology incorporating Knowledge Discovery from an existing proprietary online dating database was adopted to provide the foundation for this research (PiatetskyShapiro 1996).