Libraries are adopting Linked Open Data (LOD) and Collections as Data (CaD) approaches to present their collections as datasets for direct computational use. However, research focused on federated and reproducible access to these datasets is limited. This work aims to develop a federated and reproducible approach for extracting CaD from LOD repositories. In this context, data extracted from the single authors Jorge Juan y Santacilia and María de Zayas y Sotomayor, as well as from multiple authors from the Spanish Golden Age movement (1492–1659), are used as examples. Federated and reproducible queries are conducted using the Wikidata SPARQL public endpoint and three institutional LOD repositories on Jupyter Notebooks. The data are exported in a format compatible with computational tools (e.g., CSV) by focusing on works of a single author or works from a specific movement. Additionally, the work allows for the visualization of the queries. The results of this work provide a valuable framework for both digital humanities researchers working on datasets and libraries aiming to present their collections as accessible data for computational analysis.