Chytrá doporučení a vzdělávací cesta v rámci katalogu kurzů Skills Hub

Abstract

This thesis focuses on improving the Skills Hub platform by introducing two modules: smart mapping of courses to skills and personalized learning paths. The first module utilizes natural language processing technologies for automated and optimized skill-to-course recommendations, replacing the current manual approach. The second module allows users to select a target job role, assess their current skills, set preferences, and then the system generates an optimal learning path that combines various courses using both standard and biologically inspired methods, such as a genetic algorithm. The thesis includes a technically deep description of the selected technologies, a detailed design, and an implementation description of both modules, including the testing of the quality of the achieved results.

Description

Subject(s)

Learning platforms, Recommendation systems, Smart recommendation, Natural language processing, Vector representations, Vector database, Embedding model, Personalized learning paths, Learning path generation, Genetic algorithm, UI, Java, PrimeFaces, Python, Faiss

Citation