Supporting the detection and evaluation of disinformation using language models | seminar by PJATK research team

Prof. Adam Wierzbicki, PhD, Pavel Savov, PhD, and Witold Sosnowski, MA, from the Polish-Japanese Academy of Information Technology (PJATK) will present the results of research work over the past three years on automatic classification of WWW articles containing disinformation, classification of news articles that may contain disinformation, and data mining to detect disinformation narratives. They will present a dataset in Polish containing more than 15,000 WWW articles, including more than 5,000 disinformation articles, in 10 topics. The dataset is publicly available and will be described in the Empirical Methods in Natural Language Processing (EMNLP’2024) conference proceedings. Our guests will present new research results based on an experiment evaluating the novelty of articles that may contain disinformation, and results on the detection of disinformation narratives based on European Digital Media Observatory articles.

Prof. Adam Wierzbicki, PhD, is a full professor of computer science and a researcher with interdisciplinary knowledge and expertise in computer science, psychology and sociology. He is employed at the Polish-Japanese Academy of Information Technology, where he is vice-rector for development, director of the ICT&Design Doctoral School, and head of the Social Informatics research group. His research interests include social informatics, a field of computer science that aims to design information systems and algorithms taking into account their social and psychological impact, as well as the reciprocal influence of human behavior on information systems. Prof. Wierzbicki has pioneered research on the reliability of online content. Between 2013 and 2016, he led the Reconcile project, which investigated methods for assessing the trustworthiness of web content; this was some of the first research in this field. He is the author of the monograph “Web Content Credibility” (Springer, 2018), and more than 150 scientific papers published in renowned journals and international conferences. For more information: http://adamwierzbicki.pl

Dr. Pavel Savov has been affiliated with the Polish-Japanese Academy of Information Technology (PJATK) since 2015, where he received his Ph.D. in 2022 in the field of Technical Informatics and Telecommunications. He currently works at PJATK as an assistant professor. In his academic work, he uses natural language processing and machine learning methods in content reliability research. He also lectures on cloud computing technologies and platforms. Since 2004, he has worked in industry as a software engineer, and in recent years has been developing Big Data systems and designing natural language processing systems.

Witold Sosnowski, M.Sc., is a student at the Doctoral School of ICT&Design at the Polish-Japanese Academy of Information Technology, where he is entering his third year of study. His research focuses on the analysis of disinformation in textual data, with particular emphasis on the detection of disinformation narratives and their use in the process of identifying disinformation. He has been working professionally in artificial intelligence for four years, specializing in natural language processing. He currently works at a company developing a streaming platform with audiobooks and podcasts, where he is developing a search engine based on large and small language models.