Multidimentional Phishing Susceptibility Pridiction Model for Effective Mitigation of Potential Phishing Attack

In recent years, most of our daily activities, such as online banking and online shopping, are increasingly linked to theInternet, which makes our lives more manageable and convenient wherever we are. However, these services come with someserious security risk that can cost Internet users dearly. Experts agree that phishing is one of the critical issues that have yet tobe resolved, especially on its prediction accuracy. This is due to the fact that phishing is directed toward people instead ofmachines. The MPSPM phishing prediction model developed by Rundong Yang et al. in 2022 has achieved prediction accuracyof 89.04% on the phishing dataset as their research did not capture cognitive processes, which is one of the five factors thatinfluence phishing susceptibility in the MPSPM model. The objective of this research is to use MPSPM prediction model andmake improvement on the prediction accuracy of the model by including cognitive processes as an input into the MPSPMmodel. The results from the study indicate that identifiable models can accurately predict potential phishing victims, with theMPSPM model achieving a correct detection rate of 93.68% on the phishing dataset.