KInIT
Intelligent Technologies
Mária Bieliková
Founder & CEOMaria is inflamed in passionately building the environment for rising experts through excellent research strongly connected to the industry.
Courage, honesty and passion together with entrepreneurial spirit are the main values she adheres to and encourages. Maria developed the long-term vision and strategy for the Kempelen Institute of Intelligent Technologies.
She also conducts research focusing on human-computer interaction analysis, user modelling and personalization. She is active in discussions on trustworthy AI at the national and European levels.
Before her work at KInIT, Maria was employed at the Slovak University of Technology in Bratislava for more than 30 years — 15 as a full professor, a lead of the PeWe research group (Personalized Web) and director of the User eXperience and Interaction research centre.
She is a former dean of the Faculty of Informatics and Information Technologies (FIIT STU). She mentored many talented students and successfully participated in various international competitions.
Independent research institute
Kempelen Institute of Intelligent Technologies (KInIT) was founded in 2020 as a non-profit independent institute focused on application-oriented basic research into intelligent technologies in cooperation with the private and public sectors. It covers artificial intelligence and several areas of computing with overreach into other disciplines.
KInIT also deals with questions of ethics and human values in information technologies, specifically, in artificial intelligence. As an independent institute, KInIT aspires to articulate problems and support decision making based on scientific arguments and contribute to the Central European hub of world quality, responsible research into intelligent technology, employing top-class researchers from around the world. It is the first institute of its kind in Slovakia.
PROJECTS
YAudit
The project YAudit focuses on tracking false information spreading tendencies of recommendation algorithms on social media platforms.
Although much of the false information spreading can be attributed directly to malicious activities of users, the algorithmic preference for exciting or emotional content (regardless of its truthfulness ) is also to blame. In our research, we analyze the big platforms using so-called black-box audits. In them, we deploy groups of bots to pose as users of a social media platform.
Then, we observe what content is recommended to the bots (under a variety of conditions) and seek false information among it. This way, we are able to reliably and independently quantify the amount of false information content being spread. In a recent study, which we did over YouTube, we have shown that YouTube has (unfortunately) not improved its recommendation policies and spreads more-less the same amount of false information content as one and half year before.