Creating learning resources is a time-consuming task for teachers, as they have to find relevant information, tailor content to suit their students’ educational level, design varied learning activities, and more. To ease this workload, teachers could leverage Linked Open Data (LOD). Typically, LOD repositories contain accurate and reliable data, often linked across repositories from different organizations. However, handling this data is not easy, as it requires knowledge of Semantic Web technologies. Thanks to the emergence of large language models (LLM), the barrier to LOD that teachers must overcome can be reduced. This study aims to explore whether teachers perceive the use of LLMs to leverage LOD in the generation of ubiquitous learning tasks as helpful. For this purpose, we conducted a pilot experience with 11 high school teachers. With the results obtained in this study, we have taken initial steps toward understanding how to train teachers in the domain of artificial intelligence to effectively combine Linked Open Data and Large Language Models.