Implementasi Data Mining dengan Algoritma Naïve Bayes untuk Profiling Korban Penipuan Online di Indonesia

Abstract

Profiling of victims of crime is intended to facilitate the targeting of information dissemination and carry out prevention efforts. Profiling is helpful to increase the awareness of internet users against cybercrime. This study aims to create a sociodemographic profile based on online fraud victims using Instant Messengers in Indonesia based on the sociodemography of online fraud victims, namely age, gender, education level, domicile, occupation, duration of using the internet in a day, and Instant Messenger media used. The method used in this research is the descriptive statistical method, namely Data Mining using the snowball sampling method by sharing a link via WhatsApp. Participants were given a link to fill out several survey questions about the sociodemographic of the victim, such as age, gender, occupation, domicile, and online fraud that had been experienced through the IM application. The survey was created using GoogleForms and sent online via WhatsApp to participants who had been victims of online fraud. The Data Mining technique was used to analyze the responses of 1910 participants and then classified using the Naïve Bayes Algorithm. The results showed that the Naïve Bayes Algorithm has an accuracy percentage of 75.28%. The prediction model for the vulnerability of online fraud victims is a female respondent, aged 27.3 years, using Instagram and WhatsApp, currently living in Central Java Province, education background is high school, and the duration of using the internet more than eight hours a day, and status as a Student/College Student.