Copyright, Computation, and Literary Space: A Linked Open Data Pipeline for Contemporary Authors

This poster addresses a critical gap in Digital Humanities: how to model, analyze, and visualize twentieth-century literary works that remain under copyright while adhering to FAIR and open-access principles. Research on contemporary authors is frequently constrained by legal restrictions that limit the consultation, reproduction, and computational processing of primary texts. Rather than treating copyright as a mere obstacle, the project reframes it as a methodological stimulus – an opportunity to investigate how knowledge about an author’s life, oeuvre, and creative processes can be constructed through open, interoperable, and legally accessible data. The proposed framework integrates two complementary initiatives: OpenGadda and GaddAtlas, developed around the case study of the Italian author Carlo Emilio Gadda. OpenGadda functions both as a platform and as a workflow. Instead of engaging directly with protected texts, it relies exclusively on publicly accessible and copyright-free materials – archival inventories, bibliographies, correspondence metadata, and secondary scholarship. These data are extracted through semi-automatic procedures (regular expressions combined with manual validation), structured in tabular form, and integrated into a responsive web environment designed for exploratory access and interactive visualization. The objective is not to reproduce documents digitally, but to transform them into structured, reusable data that can be browsed, queried, and represented across multiple analytical layers. Although not yet implemented as a full RDF knowledge graph, OpenGadda already aligns its core entities (people, institutions, works, places) with external authority files such as Wikidata, VIAF, and GeoNames, laying the groundwork for a forthcoming semantic extension. GaddAtlas develops the interpretive dimension of the endeavour. Conceived as a semantic literary atlas of Quer pasticciaccio brutto de via Merulana, it reconstructs the novel’s Roman geography through close reading and manual philological extraction of spatial references. These entities are formalized in RDF via a Python-based ETL pipeline (RDFLib, YAML mappings) and structured through an ad hoc ontology grounded in OWL. The model aligns with SKOS to represent broader and conceptually nuanced categories – such as reality status and narrative role – and with PROV-O to document the provenance of interpretive assertions, ensuring transparency in modelling ambiguity and hybridity. The semantic layer is exposed through an API and visualized using advanced web-mapping and network visualization frameworks (such as MapLibre GL JS, D3.js, and Cytoscape.js), transforming the map into an analytical surface rather than a static display. By separating protected texts from structured entities and formally modelled interpretations, the project proposes a transferable, legally compliant infrastructure for computational research on contemporary literary heritage.