Security and privacy concerns in cloud-based scientific and business workflows

In science and business, workflows automate tasks, enhancing efficiency and productivity. Cloud computing further improves workflows by providing scalability and flexibility, but it also introduces security risks when sensitive data is processed in the cloud. Existing solutions primarily focus on workflow design but lack mechanisms to enforce security and respond to threats in real time.

In her thesis, Nafiseh Soveizi introduces SecFlow, a security-aware Workflow Management System (WfMS) designed to protect cloud-based workflows. SecFlow integrates real-time monitoring and adaptive response mechanisms to effectively detect and mitigate security threats.

SecFlow employs machine learning, specifically reinforcement learning (RL), to optimize security decisions. It continuously monitors cloud workflows, detects security violations, and selects cost-effective adaptation strategies to minimize disruption. The system was validated through experiments on cloud traffic and workflow datasets, demonstrating superior performance in detecting and mitigating security risks compared to traditional methods.

By ensuring security across the entire workflow lifecycle, SecFlow enhances trust in cloud-based workflow management, making cloud computing safer for both businesses and scientific research.

Dissertation: https://hdl.handle.net/11370/20880b84-b92e-4c35-ba51-a01671dcccf2