ONTOLOGY BASED EMAIL FRAUD DETECTION USING HYBRID MACHINE LEARNING
Keywords:
email; fraud detection; ml; decision forestAbstract
Emails are inevitable in this modern era. Spammers write junk/ unwanted email messages about target’s interests primarily to earn money out of it and therefore, employ creative and developed methods. Some famous techniques are employed by fraudster to commit email fraud, leading to personal information and financial losses incurred/ faced by innocenttargets. Therefore, identifying spam and email fraudis a widely researched topic. Resultantly, to overcome existing gaps and improve already developed systems, this research work presents a solution devised on the concept of ontology-based email fraud detection. Moreover, ontology-based email fraud detection is enhanced using hybrid machine learning, incorporating two class decision forest learning algorithm to detect spam/ fraudulent emails with more accuracy. The ontology based technique detect new types of fraudulent emails with 99.2% accuracy as compared to conventional techniques. Experiments performed using the proposed solution show promising results with greater accuracy.