Opis sylwetki doktoranta
Arkadiusz Modzelewski is a third-year PhD student simultaneously enrolled in doctoral programs at two universities: Brain, Mind, and Computer Science PhD program at Università degli Studi di Padova and Computer Science PhD program at the Polish-Japanese Academy of Information Technology. His enrollment in both programs is possible thanks to the Cotutelle agreement, which allows him to make research more internationally recognizable and prestigious. His research, supervised by Prof. Adam Wierzbicki and Prof. Giovanni Da San Martino, focuses on detecting disinformation, manipulation, and persuasion techniques in textual data, bridging technical expertise with societal impact. Arkadiusz has presented his work at international conferences, such as the top-tier Conference on Empirical Methods in Natural Language Processing. Arkadiusz participated in international workshops like SemEval and CheckThat! and achieved successful results.
Opis tematyki projektu badawczego
Mitigating the spread of disinformation on the internet has become a critical societal challenge. Numerous significant events, including the COVID-19 pandemic, highlight disinformation’s negative impact on individuals and society. Disinformation often involves manipulative and persuasive techniques, making it essential to address not only the detection of false information but also the identification of manipulative tactics. This research addresses the complex challenge of identifying disinformation and persuasive techniques in news articles and other online texts. Although the study encompasses multilingual texts, the focus is on the Polish language, aiming to support less-researched language. The project will contribute to the development of innovative datasets and Artificial Intelligence tools for detecting disinformation and persuasive tactics. It will be conducted under a Cotutelle Agreement between two universities: the University of Padua and the Polish-Japanese Academy of Information Technology.
Wybrane publikacje
Modzelewski, Arkadiusz, Giovanni Da San Martino, Pavel Savov, Magdalena Wilczyńska, and Adam Wierzbicki. „MIPD: Exploring Manipulation and Intention In a Novel Corpus of Polish Disinformation.” In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pp. 19769-19785. 2024.
Sosnowski, Witold, Arkadiusz Modzelewski, Kinga Skorupska, Jahna Otterbacher, and Adam Wierzbicki. „EU DisinfoTest: a Benchmark for Evaluating Language Models’ Ability to Detect Disinformation Narratives.” In Findings of the Association for Computational Linguistics: EMNLP 2024, pp. 14702-14723. 2024.
Modzelewski, Arkadiusz, Witold Sosnowski, Magdalena Wilczynska, and Adam Wierzbicki. „Dshacker at semeval-2023 task 3: Genres and persuasion techniques detection with multilingual data augmentation through machine translation and text generation.” In Proceedings of the 17th International Workshop on Semantic Evaluation (SemEval-2023), pp. 1582-1591. 2023.