Data Science & Explainable AI
The team undertakes both empirical and design research, with emphasis on explainability aspects of artificial intelligence. The empirical research primarily covers psychological aspects and biases in machine learning, both wrt. training data and result interpretation. The design of new algorithms, in turn, focuses on rule discovery, both in propositional and relational setting, and on the methods of mechanizing hypothesis formation, which integrate data mining with formal logic (observational calculi). Knowledge graphs are also considered both as source and/or target of machine learning.
Besides inductive methods, deductive approaches are also applied on ontologies and knowledge graphs, via structural design patterns. Finally, a part of the group investigates selected topics of artificial general intelligence, namely, measuring the intelligence of artificial agents through comprehensive tests.
Main directions of the research:
-
Explainable machine learning and knowledge discovery
-
Knowledge representation, ontologies and knowledge graphs
-
Measuring the intelligence of artificial agents
#Artificial intelligence, #Machine learning, #Knowledge discovery, #Knowledge graphs, #Semantic web
Representatives of the research team
Petr Berka
Tomáš Kliegr
Jan Rauch
Vojtěch Svátek
Ondřej Zamazal
Top publications of the team
-
KLIEGR, Tomáš, BAHNÍK, Štěpán, FÜRNKRANZ, Johannes. A review of possible effects of cognitive biases on interpretation of rule-based machine learning models. Artificial Intelligence. 2021, Vol. 295.
-
MÁŠA, Petr, RAUCH, Jan. A novel algorithm weighting different importance of classes in enhanced association rules. Knowledge-Based Systems, 2024, Vol. 294.
-
SVÁTEK, Vojtěch, ZAMAZAL, Ondřej, NGUYEN, Viet Bach, IVÁNEK, Jiří, KĽUKA, Ján, VACURA, Miroslav. Focused categorization power of ontologies: General framework and study on simple existential concept expressions. Semantic Web, 2023, Vol. 14.
Selected research projects
-
EU HE Onto-DESIDE – Ontology-based Decentralized Sharing of Industry Data in the European Circular Economy, 2022-2025
-
EU COST Action GOBLIN: Global Network on Large-Scale, Cross-domain and Multilingual Open Knowledge Graphs, 2024-2028
-
IGS VSE “Evaluation of general intelligence of artificial agents”, 2023-2025