Mark David Longar, Erik Novak and Marko Grobelnik<\/p>\n
Abstract<\/strong> \u00a0<\/p>\n\n\n\n
A key limitation of state-of-the-art large language models is their
lack of a consistent world model, which hinders their ability to
perform unseen multi-hop reasoning tasks. This paper addresses
this by extracting local world models from text into a system-
atic first-order logic framework, enabling structured reasoning.
Focusing on the educational domain, we present a multi-step
approach using Prolog to represent and reason with these mod-
els. Our method involves segmenting educational texts, generat-
ing Prolog definitions, and merging them into a comprehensive
knowledge graph. We successfully extracted several small models
and manually verified their accuracy, demonstrating the poten-
tial of this approach. While promising, our results are currently
limited to small-scale models.<\/p>\n