The employees of Radar Cyber Security come from 21 nations. Fatemeh Amiri works at Radar Cyber Security’s Cyber Defense Center.

Thank you Fatemeh for taking the time to talk to us. How did you become aware of Radar Cyber Security?

I come from Iran. After graduating with my bachelor’s degree in computer science and my master’s degree in information technology I moved to Austria to continue my studies. My research topic was about privacy protection using machine learning. I hoped to refine and close privacy gaps by presenting new methods with embedded privacy-preserving algorithms which are essential for all data mining tasks, cloud-based services, e-business applications and IoT structures.

What makes working at Radar Cyber Security special and exciting for you?

The combination of my research topic combined with practical work in the IT security field. I am a security analyst and a PhD candidate at the University of Vienna. This means I’m close to the practical side of my research and can exchange ideas and work closely with developers and a team of internal researchers. This gives me the opportunity to share ideas and discuss concepts and methods in a huge community of highly specialised experts.

How important is the exchange with colleagues for you?

The exchange is very valuable for me, because my research topic on privacy is in a scarce research field. My research findings show that traditional methods are not sufficient to solve big data problems. New methods like machine learning seem to be the smarter solution to achieve better results.

What is the main goal you want to achieve with your research?

My primary objective is to improve privacy while keeping the efficiency and accuracy of data mining tasks that handle the operation and considering other factors which also influence the overall process of the system.

What is your approach to more data security?

Essentially, I try to anonymize the respective sensitive data with my model. In order to minimize the information loss, machine learning methods like genetic algorithms and fuzzy sets are used to hide this data the best way possible and to keep a balance between all the defined goals. First results have proven the efficiency of the proposed methods. Improving the accuracy of the final data mining task is the next step. The aim of this approach is to apply it to other data mining tasks and break it down to a simple formula to implement it in various fields.