Matthieu Jimenez

Matthieu Jimenez is a Research Fellow at the University of Luxembourg's Faculty of Science, Technology and Medicine, working within the Serval team at the Interdisciplinary Center of Security and Trust under Professor Yves Le Traon. Since Summer 2020, he has focused on software vulnerability analysis and detection through machine learning approaches while teaching Introduction to Computer Science and contributing to Software Engineering II and Software Testing courses for Computer Science Bachelors.

Education:

• PhD in Computer Science (2018) from University of Luxembourg with thesis 'Evaluating Vulnerability Prediction Models' (excellent grade)
• Engineer's degree (MSc equivalent) in Computer Science with major in Information Security from Polytech'Nice Sophia (2014)
• Exchange year at Polytechnique Montréal (2012-2013)

Jimenez's research centers on applying machine learning to software security challenges, particularly vulnerability prediction modeling and source code analysis. His work bridges theoretical machine learning concepts with practical security applications, developing tools that help identify vulnerabilities in software systems through advanced data analysis techniques. He has made significant contributions to understanding how code naturalness metrics can reveal security flaws and how to properly evaluate prediction models in real-world scenarios.

His publication record shows a consistent focus on vulnerability prediction methodology, with particular attention to practical implementation challenges like handling noisy historical data and proper model evaluation. The progression from his PhD work on evaluation frameworks to recent applications in real-world prediction demonstrates his commitment to solving practical security problems through rigorous research.

Scientific Awards:

• Distinguished paper award at FSE'19 for 'The importance of accounting for real-world labelling when predicting software vulnerabilities'

Jimenez serves on program committees for major software engineering conferences including ECOOP, ASE, and ICSE, demonstrating his standing in the research community. He has advised on artifact evaluation processes and contributed to research paper review committees. His work with the Serval team at SnT involves developing practical security tools while advancing theoretical understanding of vulnerability prediction.

He leads development of the tuna project (Tuning Naturalness Analysis) for source code analysis and the data7 tool for automatically generating vulnerability datasets, both of which have become valuable resources for the software security research community. These tools enable more rigorous evaluation of vulnerability prediction models by providing standardized datasets and analysis frameworks.