Towards Machine Learning Models as a Key Mean to Train and Optimize Multi-view Web Services Proxy Security Layer

Anass Misbah, Ahmed Ettalbi

Abstract


Muti-view Web services have brought many advantages regarding the early abstraction of end users needs and constraints. Thus, security has been positively impacted by this paradigm, particularly, within Web services applications area, and then Multi-view Web services.

In our previous work, we introduce the concept of Multi-view Web services to Internet of Things architecture within a Cloud infrastructure by proposing a Proxy Security Layer which consists of Multi-view Web services allowing the identification and categorizing of all interacting IoT objects and applications so as to increase the level of security and improve the control of transactions.

Besides, Artificial Intelligence and especially Machine Learning are growing fast and are making it possible to simulate human being intelligence in many domains; consequently, it is more and more possible to process automatically a large amount of data in order to make decision, bring new insights or even detect new threats / opportunities that we were not able to detect before by simple human means.

In this work, we are bringing together the power of the Machine Learning models and The Multi-view Web services Proxy Security Layer so as to verify permanently the consistency of the access rules, detect the suspicious intrusions, update the policy and also optimize the Multi-view Web services for a better performance of the whole Internet of Things architecture.

Keywords


Cyber Security; Internet of Things; Cloud; Multi-view Web services; Security layer; WADL; WSDL; Restful Architecture; Artificial Intelligence; Data science; Machine learning

Full Text:

PDF



International Journal of Recent Contributions from Engineering, Science & IT (iJES). eISSN: 2197-8581
Creative Commons License
Indexing:
DBLP logo DOAJ logo