The increasing integration of technology into language education presents new opportunities for personalised learning, yet developing and maintaining AI-powered tools like chatbots often requires significant technical expertise. This thesis addresses the gap in practical methodologies and accessible tools for language teachers in specialised domains, specifically focusing on the Swiss public transportation sector. It investigates how a knowledge graph-based low-code application can enhance the development of educational chatbots for these teachers.

Employing a design science research approach, the study involved the development of a knowledge graph and a corresponding low-code teacher interface using Metaphactory, HTML, and CSS, grounded in a refined ontology development methodology. Evaluation revealed a high success rate in integrating diverse language teaching materials into the knowledge graph, demonstrating its robustness in capturing specialized domain knowledge. Prototype testing yielded positive teacher feedback, confirming the intuitive design and usability of the low-code interface in abstracting complex data management. Crucially, the knowledge graph's structure was validated to support essential chatbot functionalities via an accessible platform.

The findings underscore the utility of a knowledge graph-enhanced low-code application in empowering language teachers. By reducing technical barriers and accelerating content development, this work contributes to bridging the gap between pedagogical expertise and technological implementation. It paves the way for more adaptable, personalised, and effective language learning experiences in niche professional contexts, fostering greater autonomy for educators in the digital age.