Upcoming Open Mic - Using LLMs to build glossaries of Czech legislation
On Friday 12 December 2025 at 10:30 Martin Ledvinka will talk about his experiments with extracting glossaries of Czech legislative documents using LLMs. You can join us at this link.
Abstract
For several years, there has been an effort to build a knowledge base of Czech legislation (and similar normative documents). This knowledge base consists of glossaries of terms used in the documents and, in some cases, also ontological models representing deeper relationships between the terms. However, creation of such artifacts has been purely manual work, requiring a lot of time and expertise.
In this talk, I will present my experiments with using LLMs to extract glossaries from Czech legislative documents. I will discuss the approach, sketch the pipeline that could be used for a large-scale application, and present initial results of the experiments.
Related links: