News and events archive

From the faculty

  • Title image

    Czech scientist in the heart of Paris: Dr. Švábenský on his internship at Sorbonne Université

    In 2022, we informed you about the Joseph Fourier Prize in the field of Computer Science and Informatics, which was awarded to RNDr. Valdemar Švábenský, Ph.D., from FI MUNI. As the winner, he had the opportunity to spend a month at a French university. He eventually spent this month at Sorbonne Université in September 2023, and now he shares with us his impressions and experiences from this internship.

    What was the main goal of your internship at Sorbonne Université and how did it relate to your winning work "Automated Feedback for Cybersecurity Training"? The primary goals were to establish international cooperation, develop joint research with the host lab, and get a taste of French culture. Moreover, during that month, we developed a conference paper and a poster, which we are completing remotely in the following months. The internship mainly focused on the analysis of educational data, which ties into my final thesis.

    What projects did you work on during the internship, and with whom? I visited the MOCAH research group: Models and Tools in Knowledge Engineering for Human Apprenticeship. They are dedicated to the domain of learning analytics, researching methods like machine learning for automatic information extraction from data originating in the context of education, aiming to improve some aspect of teaching. At FI MUNI, this topic has long been addressed by Radek Pelánek, and in the CYBERSEC lab we approach it through practical cybersecurity training.

    My host, François Bouchet, has several projects in the field of learning analytics at Sorbonne. We chose to focus on detecting urgent student posts in discussion forums of university courses. In large online courses, students ask many questions, and it's challenging for a teacher to monitor everything and respond in time. We prototyped models that aim to automatically determine which post requires a quick reaction from the teacher. I have an article on this topic from my year-long stay in the USA. Together with Dr. Bouchet, we took it a step further, and it resulted in a poster.

    The article I developed there addresses a very current topic of algorithmic bias, i.e., measuring and evaluating systematic errors in machine learning models applied in the context of education. If a model with such systematic error is deployed in educational practice, it can lead to incorrect or even unfair decisions. Therefore, it is appropriate to check and evaluate models against bias metrics. Mélina Verger, a member of the MOCAH lab, is also focused on this topic for her dissertation, making this a mutually beneficial collaboration.

    Did you choose the university you visited, or was it recommended to you? And why did you ultimately choose Sorbonne Université? 

    I chose it myself. I started by reviewing the proceedings of key conferences in learning analytics over recent years and looked at who was publishing from France. I went through the profiles of these people on Google Scholar to see what they had been focusing on and narrowed it down to about 3-4 universities. Then I asked my then colleagues at the University of Pennsylvania if they could recommend someone from my list. A colleague connected me with Dr. Bouchet, whom she had met at one of those conferences.

    What were your first impressions of Sorbonne Université, and how did they differ from your expectations? 

    My impressions were very pleasant; I had everything I needed from the first day, so I lacked nothing. I was positively surprised that everyone was very willing to speak with me in English, as unfortunately, I do not know French. I somewhat expected the stereotype that the French do not like to speak English, but everywhere in Paris, they were accustomed to tourists and English, so it was no problem.

    Congratulations on your pleasant experience communicating in English in Paris. Could you tell us how your stay at Sorbonne Université has reflected in the progress of your research work? Is there anything from this internship that you apply, or will apply, in your further work at FI MUNI?

    Surprisingly, everything went quite smoothly. The work did not get stuck anywhere, we had regular checkpoints. The agency overseeing the internship took care of almost all the administration and accommodation, so I was well taken care of. As for free time, I'm used to traveling a lot and handling things abroad, often alone, so I secured all life necessities and leisure activities myself.

    The challenge was sometimes getting somewhere with the Parisian metro. At certain hours it was so full that it was impossible to board, and it was necessary to just wait for several more connections until one of them was free enough to squeeze in. One must account for this or use another mode of transport.

    We are glad that you managed well and that you always, even by metro, reached your destination. What experiences in Paris were most inspiring for you?

    When I travel, I enjoy meeting local people and discovering new places. There are many beautiful locations in Paris, and I have nice memories from there. An evening walk around the illuminated Eiffel Tower or along the banks of the Seine is breathtaking.

    What message do you have for students and colleagues at FI MUNI who are interested in similar international internships?

    I 100% recommend everyone to go for it if they have the opportunity. Ever since I was (during my master's studies) on an Erasmus+ stay in Denmark, I have been a proponent of such international trips. It's a great experience. You will see a different part of the world, something there will appeal to you more and you might try to bring it home, something will be better at home and you will start to appreciate it more.

    Thank you, and we wish you success in your research and future internships, and we look forward to sharing more experiences from your travels.

    Web address
    Original bulletin in the Information system.