Deep Learning-Based Framework for Detecting Malicious Insider-Inspired Cyberattacks Activities in Organisations.

Cyberattacks are happening at an alarming rate both in developed and developing countries. This is due to more users now being connected to the global village (internet). Significant strides have been taken by organisations to protect information technology assets together with data, by doing defense-in-depth, using firewalls and access control approaches collectively. These approaches work well in detecting attacks by outsider cyber-attackers. In recent cyberattacks the perpetrators have been those within the organisation, as they can easily bypass security measures especially those with high privileges and they can go undetected for quite a long time. We propose a deep learning approach termed Automatic_ IDS_ Deep model (framework) that is infused with intrusion detection systems to give timely detection of malicious activities by those within the organisation. Experiments were conducted and averaging of results was done to determine accuracy, recall, and precision of the proposed model. The model (framework) offers better results on its performance in detecting attacks that are perpetrated within the organisation. [ABSTRACT FROM AUTHOR]