The Laboratory for Semantic Analysis (LSA) conducts research on natural language as a unified whole using methods from computer science and applied mathematics. We examine language as a collection of vectors in a semantic space, which enables broad theoretical analysis of language using methods from complex systems theory, topological data analysis, and nonlinear equations analysis.

Practical applications include the development of new approaches for building language models in manifold learning and large language models. Among the projects with high commercialisation potential are ‘Catch the Bot,’ ‘Jokes Aside,’ and ‘Tree of Knowledge.’

The laboratory is inherently interdisciplinary and collaborates with other faculties and laboratories. It aims to create a large-scale model of natural language, which allows for the development of new approaches to building large language models and semantic technologies.