About two months ago, I had the privilege to attend the 2025 Annual Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics (NAACL), which took place in Albuquerque, New Mexico, on April 29–May 4. This was my third big international in-person conference and second ACL experience, as I previously went to EAMT 2023 (the 24th Annual Conference of the European Association for Machine Translation) and EACL 2024 (the 18th Conference of the European Chapter
of the Association for Computational Linguistics) during my Master’s studies. But it was the first time I would present a paper at such an event, making NAACL 2025 a crucial step in my academic career.
What made it even more meaningful was the fact that my paper was accepted at the 9th Joint SIGHUM Workshop on Computational Linguistics for Cultural Heritage, Social Sciences, Humanities and Literature (LaTeCH-CLfL 2025), co-located with NAACL. Last year, when I attended EACL in Malta, I felt somewhat overwhelmed at the beginning of the conference. As someone coming from a translation background and only starting to learn programming, I found most of the talks too technical and difficult to understand. At the LaTeCH Workshop, however, which took place on the last day of EACL, I felt much more in my element: the topics discussed there focused on the humanities rather than the computational aspect, which aligned closer with my interests. Little did I know that the workshop organizer, Stefania Degaetano-Ortlieb, would later become my supervisor during my PhD.
This year, after wrapping up the exposé for my PhD project under the guidance of my supervisor, I decided to submit it as a position paper to LaTeCH 2025. To my pleasant surprise, it was accepted. In this work, titled Interpretable Models for Detecting Linguistic Variation in Russian Media: Towards Unveiling Propagandistic Strategies during the Russo-Ukrainian War, we discuss the theoretical background of propaganda detection in the context of the Russo-Ukrainian War and propose a robust methodological framework to study linguistic variation and change that might signal propagandistic strategies across two dimensions: text types and time. I opted to present my paper as a poster, as I prefer this format over talks due to its interactivity. It turned out to be a good decision: apart from sharing my work with others, I gained valuable feedback and had insightful conversations with the conference participants who came up to my poster.
As for the main conference, which lasted from April 30 till May 2, its range of topics covered all areas of computational linguistics: from machine translation, multilinguality, and computational social science to information retrieval, machine learning, and language generation, with a special theme called “NLP in a Multicultural World”. The conference included three keynote talks and numerous oral presentations and posters. Workshops and tutorials were held on the last two days, May 3 and 4. Apart from LaTeCH, I attended some talks at other workshops, such as NLP4DH (the 5th International Conference on Natural Language Processing for Digital Humanities) and MWE (the 21st Workshop on Multiword Expressions). It was inspiring to see such a wide variety of NLP topics explored in so many creative ways and from different perspectives by researchers with diverse backgrounds. Not only listening to presentations, but also talking to others during coffee breaks and social events filled me with ideas for my own project.
The social aspect of the conference also played a crucial role in making it memorable. Although we were a small group from the Language Science and Technology Department at Saarland University — consisting of me, my supervisor, a postdoc colleague, and a Master’s student — it was great to share this experience with them and the new friends I made at NAACL. It was incredible to see how “small” the NLP community actually is, despite the huge number of people attending the conference. I’ve met many people whom I saw previously at other events, or read their work, or who turned out to be former colleagues of my current co-workers. The welcome reception and social dinner organized at NAACL were excellent opportunities to converse and exchange ideas in a less formal environment. And of course, in our free time, we also explored the surrounding nature, such as the Petroglyph National Monument, where we could see petroglyph images carved by Native American tribes and early Spanish settlers in a canyon nestled among some truly stunning mountains. Moreover, we tried the delicious local cuisine, influenced by Mexican, Native American, and Spanish culinary traditions.
I came home feeling inspired and motivated to continue my research. Once again, I was reminded that conferences are a great space to share knowledge, expand your network, establish collaborations, and — no less important — travel, discover new places, and just have fun.
Anastasiia Vestel completed her MA at an Erasmus Mundus program called “European Master’s in Technology for Translation and Interpreting” (EM TTI), having studied at Ghent University, Belgium, and the University of Málaga, Spain. As an EM TTI student, she had the opportunity to acquire practical skills in the field of computational linguistics and natural language processing, which eventually led her to apply for a PhD at the CASCADE (MSCA) Doctoral Networks. Her research interests include literary machine translation, diachronic semantic change, as well as propaganda and disinformation detection.
CASCADE is a collaboration between University College Cork, University of Sheffield, University of Helsinki, Katholieke Universiteit Leuven, and Universität des Saarlandes. Funded by Horizon Europe under the Marie Skłodowska-Curie Actions (MSCA) Doctoral Networks and the UKRO.