Design and development of Malaysian cybersecurity profiling framework: Towards creating a recommendation system to combat cybercrime

Yeo, Han Rong (2022) Design and development of Malaysian cybersecurity profiling framework: Towards creating a recommendation system to combat cybercrime. Master dissertation/thesis, UTAR.

PDF Download (1426Kb)

Abstract

Malaysia is deeply endangered by cybercrime and suffer huge economic losses. To against it, Malaysia government established institutions like MyCERT and CyberSAFE trying to increase the cyber security awareness of public by providing cyber security advice on the website but it not working as expected due to the ambiguity and complexity of the security advice. Moreover, evaluation of the cyber security awareness required a measurement method. Currently, most of the measurement method are using questionnaire and there is no measurement method that is applicable to an individual. This result in the awareness of public is unmeasurable. This research proposed a primary cybersecurity profiling framework for Malaysian by integrating Malaysia cybercrime situation, auto cyber security awareness measurement, and collaborative recommendation system. The users are first categorized into 3 levels based on their awareness result. The system will then provide recommendation to the users based on their awareness level. The profiling system is able to profile the users automatically using the data collected from users’ computer. Using the matrix factorization collaborative filtering together with RNN-LSTM to train the model, the model with ISA-aware feature improved the RMSE by 0.12 and accuracy by 12.2%. Furthermore, the overall of cyber security awareness of users is improved after using the system.

Actions (login required)