Learning from the Best

Congratulations to BIFOLD researcher Anne Josiane Kouam, who was selected as one of the most promising young researchers to participate in the 11th Heidelberg Laureate Forum (HLF) 2024, which will take place from September 22 to 27, 2024, in Heidelberg, Germany. “It is a great honor to have been selected to join such a distinguished panel and I look very much forward to engaging with and drawing inspiration from the foremost researchers in my field”, says Anne Josiane Kouam, who was also awarded an Abbe Grant 2024. The Abbe Grant is issued from the Carl-Zeiss-Stiftung, an academic partner of the Heidelberg Laureate Forum Foundation (HLFF). Thirty young researchers have been selected to receive the grant based on diverse criteria and academic credentials.

Anne Josiane Kouam is a postdoctoral researcher at the Machine Learning and Security team under the supervision of Prof. Dr. Konrad Rieck and completed her Computer Science engineering degree at the National Advanced School of Engineering in Cameroon in 2019. Following that, she pursued a PhD from INRIA Saclay and Ecole Polytechnique, France. Focused on privacy and security in mobile and cellular networks, her research explores the evolving landscape of threats, particularly those arising from the intersection of Machine Learning and network security.
The HLF is a networking conference where 200 carefully selected promising young researchers in mathematics and computer science spend a week interacting with the most exceptional mathematicians and computer scientists of their generations, recipients of the most prestigious awards in the disciplines, such as the Abel Prize, ACM A.M. Turing Award, ACM Prize in Computing, Fields Medal, IMU Abacus Medal, and Nevanlinna Prize. Established in 2013, the HLF is annually organized by the Heidelberg Laureate Forum Foundation.

About Anne Josiane Kouam: 
Her research delves into the crucial areas of mobile security and privacy, which have become increasingly significant due to the pervasive access provided by mobile networks. As these networks connect a growing number of users and devices, the potential for attacks escalates. Her work addresses several key issues:

• Bypass fraud: This involves fraudsters redirecting international calls to appear as local calls within the destination country. Apart from causing substantial financial losses, this attack compromises users' privacy by allowing fraudsters to eavesdrop on international phone conversations.
• Phishing attacks: Traditionally associated with email communication, phishing attacks have expanded to include text messages on mobile networks, known as "smishing" attacks. This broadens their reach and makes it easier to deceive unsuspecting individuals.
• User tracking attacks: Leveraging mobile devices' pervasiveness, these attacks infer users' locations through accessible information without explicit permission.

Furthermore, the widespread reach of mobile networks ensures the comprehensiveness of mobile network datasets, such as Charging Data Records, which include users' mobile service usage and location information. While valuable for research across various domains, these datasets also pose challenges due to the inclusion of sensitive user information. Research explores the use of deep generative models to create privacy-preserving realistic mobile datasets. This includes:

• Investigating the vulnerability of released mobile datasets to de-anonymization attacks.
• Investigating privacy leaks of generative models.
• Evaluating the trustworthiness of crowd-sourced mobile network datasets.

Anne Josiane’s investigations span both offensive and defensive strategies, exploring the effectiveness of attacks, particularly when strengthened by recent advancements in machine learning. This comprehensive approach not only enhances the understanding of these threats but also paves the way for innovative, widely applicable machine learning-driven defenses.