Hey, I'm Ivan Vykopal 👋
I'm a Research Assistant and PhD Student at 
Kempelen Institute of Intelligent Technologies and member of the NLP team, focusing on machine learning, deep learning, the analysis of large language models, and computer vision.
I hold a Master's degree in Computer Science from the Faculty of Informatics and Information Technologies at the Slovak University of Technology. During my studies, I focused on artificial intelligence, deep learning, participating in several research projects related to the application of deep learning and computer vision in the field of medicine. My Master's thesis and research delved into deep neural networks in the domain of medical image processing, with a specific focus on the segmentation of higher morphological structures for identifying possible cardiac transplant rejection.
PhD Topic: Multilingual Natural Language Processing for Supporting Fact-Checking
Multilingual natural language processing (NLP) plays a critical role in designing models that can handle multiple languages simultaneously or focus on specific languages. While English dominates the field due to the abundance of available data for training powerful models, expanding NLP capabilities to other languages is essential, particularly for low-resource languages. This expansion becomes even more crucial in the context of false information, where NLP techniques can be used to detect and counter false information proliferating on the internet and social media platforms. As online spaces continue to grow as key sources of information, multilingual NLP emerges as a vital tool in the fight against misinformation, ensuring that fact-checking systems and models are equipped to address global and multilingual content effectively.
With the recent advancement in large language models (LLM), we identify an opportunity to employ their potential and extensive capabilities in various complex NLP tasks, including fact-checking. Our research aims to evaluate the current capabilities of LLMs for supporting fact-checking and propose new approaches to mitigate the spread of false information in multilingual contexts while assisting fact-checkers in their efforts.