A Prediction Model to Prevent Phishing Attacks on E-Mails Using Data Mining Techniques


Abstract:

The current study presents a design and implementation of a prediction model that prevents phishing attacks on emails using data mining techniques. Hereby, we collected appropriate data sets in order to identify the characteristics of the infected emails, which allow a phishing attack to be successful from three significant sources, being Monkey.org, Enron.org, and PhishTank.com. For the design and implementation of the model, we used the CRISP-DM methodology in order to generate the detection model based on the characteristics that recognize an email as phishing. Likewise, we developed a descriptive analysis using Decision Trees, Bayesian Networks, and Random Forest as data mining algorithms, whose most precise has been Random Forest. Once the email was detected as infected, the algorithm quarantines the email. Subsequently, it sends a notification message through WhatsApp. Finally, we evaluated the model in a controlled environment. The results demonstrate the functionality of the model since it achieved a precision score greater than 97.28% in the detection of phishing emails.

Año de publicación:

2022

Keywords:

  • Data Mining
  • pbkp_rediction model
  • random forest
  • Phishing

Fuente:

googlegoogle
scopusscopus

Tipo de documento:

Conference Object

Estado:

Acceso restringido

Áreas de conocimiento:

  • Minería de datos
  • Ciencias de la computación

Áreas temáticas de Dewey:

  • Programación informática, programas, datos, seguridad
  • Funcionamiento de bibliotecas y archivos
  • Criminología
Procesado con IAProcesado con IA

Objetivos de Desarrollo Sostenible:

  • ODS 9: Industria, innovación e infraestructura
  • ODS 17: Alianzas para lograr los objetivos
  • ODS 8: Trabajo decente y crecimiento económico
Procesado con IAProcesado con IA