Towards Securing Mobile Communication from Spyware Attacks with Artificial Intelligence Techniques

Nowadays security in mobile phones is a crucial issue, attacker attacks through messages and emails and when the user opens these malicious links, they can easily insert the malware into the user’s mobile phone and get the all information/data of the user. As the number of mobile users is growing on a higher level for multiple types of applications like emails, online transaction, messaging etc. the attacker easily get all information of the user's mobiles like bank details, user credentials, photos and videos etc. In the current situation, where attacks have increased in exponential frequency, a significant problem now is putting forward innovative technology. As machine learning becomes more prevalent, it may now provide clever solutions for the early detection of spyware attacks for different mobile platforms. As machine learning becomes more prevalent, it may now provide clever solutions for many applications including early spyware attack detection. The approach that is suggested in this research work acts as a shield for mobile devices that are receiving malicious data packets from an attacker. To ensure that only authorized data packets are transferred to mobile phones, it is recommended to use a trained machine learning chip for spyware attack detection. This paper presents the comparative analysis of three artificial intelligence techniques for early spyware detection. The dataset for training and testing artificial intelligence methodologies is taken from Canadian Institute for Cyber security data repository. The results prove that the best method for detecting spyware attacks at an early stage is Convolution Neural Network.