CYBER FRAUD PROFILING WITH ROUTINE ACTIVITY THEORY USING DATA MINING TECHNIQUES

Abstract

Cyber profiling, as one of the supporting parts of digital forensics, is not only used to record and investigate cybercriminal behaviour. It can also be used to profile victim demographics based on victim characteristics. This study aims to create a cyber-fraud pattern based on a profile created from RAT. This research is expected to be input for internet users in Indonesia, especially IM users such as Instagram, Facebook, WhatsApp, and Telegram. The data collection method in this study uses data mining technology on structured and unstructured data. Structured data was obtained by conducting data mining on the number of cases registered in district courts in Indonesia from January 2021 to January 2022, and the unstructured data was obtained from socio-demographic victims of online crimes. The analysis using the Naive Bayes Algorithm produces a predictive model, which shows the results of online fraud victim profiles based on the weights for each attribute. Cyber-fraud profiling based on RAT with Naïve Bayes Algorithm yields the following findings: Potential Offender Elements: Male, using Facebook, WhatsApp, and Instagram, and crime scene region in Special Capital Region of Jakarta; Elements Suitable Target: Female, using Instagram, WhatsApp, and Facebook, living in the Special Region of Yogyakarta, spending time on the internet more than 8 hours a day, and have more than three IM applications; and Guardianship: Lack of knowledge about Cyber Fraud.

Keywords

• cyber fraud
• cyber profiling
• data mining
• routine activity theory