This study proposes a model for personalising AI assistants for specialised translation through prompt engineering, context adaptation, and the integration of terminological resources. The resulting model incorporates an agent profile, an educational context modelling layer, a contextual framework, and a structured workflow. This model was validated using Agent Builder in Microsoft Copilot Studio, as well as proto-agents in ChatGPT and Claude, to translate Bulgarian educational texts related to computer science and educational technology. The results suggest that personalised AI assistants can facilitate the creation of terminologically consistent and pedagogically appropriate translations, even under constraints of limited time and technical resources.
Paper
The full text of this publication is not hosted on 44B due to licensing.
Read it at OpenAlex