Abstract
Quality scientific inquiries depend on access to data distributed over the entire globe. Linked open data (LOD) and FAIRness play major roles in ensuring access to data that scientists need to answer interesting questions. However, a data model and a query language to compute responses to complex scientific inquiries remain outstanding. As the recent emergence of large language models (LLM) reshape how we interact with machines, an intriguing prospect of posing scientific inquiries to smart machines suddenly appears realizable in which a natural language ChatBot is empowered with a LOD knowledgebase as its data source. In this paper, we introduce a model for an LLM interpreter, called ProAb, that aims to answer natural language scientific queries using a structured query language called Needle in which the LOD is viewed as a set of tables. We discuss the contours of ProAb, present its preliminary and experimental design, and highlight its salient features using an illustrative example. It should be apparent that a full automation of ProAb is feasible with further research.