Naples Dante Project

Series: Miscellaneous

Language: Italian/English

Abstract: The Naples Dante Project (NDP) is a digital infrastructure for the study of the tradition of Dante’s Commedia and its exegetical and iconographic apparatuses, founded and directed by Gennaro Ferrante. The project is conceived as an online aggregator of interoperable databases devoted to the manuscript, printed, and iconographic tradition of the poem, developed according to semantic standards (CIDOC-CRM/LRMoo) and interoperability protocols (IIIF). The project is grounded in the awareness that the tradition of the Commedia represents a limit case for computational approaches, due to the scale of the corpus (around six hundred manuscripts, more than one hundred and thirty fragments, and numerous printed editions) and its intrinsically integrated nature, in which material support, text, language, images, and exegetical apparatuses are inseparable. The NDP_CADMUS database brings together the main corpora of the tradition: illuminated manuscripts (IDP_Illuminated Dante Project), fragments (FraC-Fragments of Commedia), printed editions (DIP_Dante in Print), and graphic cycles (DaD_Dante Drawings). These are complemented by Dante ICON, a transversal module for the indexing and comparative analysis of iconographic content, and by the DCT_Dante Critical Texts corpus, which gathers and aligns the major scholarly editions of the Commedia from 1862 onwards, making them searchable as data. Data modelling is implemented through the modular environment Cadmus and semantically represented in the ALIGHIERoo ontology, which formalises the relationships among textual, codicological, and iconographic entities and enables integration with other standards and systems. The data derive both from expert autoptic analysis of the physical supports and from automated processes based on Machine Learning models. To support the entire data lifecycle, NDP adopts the pipeline developed within the MAPP (Manuscripts Philological Pipeline) project, which integrates tools for metadata creation (Cadmus), automatic transcription and text alignment (eScriptorium), semantic data management (Virtuoso), and workflow orchestration (MetaFAD), enabling the acquisition, processing, analysis, and online publication of resources within an interoperable environment compliant with FAIR principles. Within this framework, the project integrates Handwritten Text Recognition (HTR) techniques, already operational within the pipeline, for the semi-automatic transcription of manuscripts, while additional Machine Learning-based components are under development for automatic handwriting identification, large-scale linguistic analysis for the historical and geographical positioning of witnesses, automatic alignment of transcriptions for detecting variants and lacunae, metrical and prosodic analysis of the poem, and Computer Vision techniques for the study of iconographic content. These tools contribute not only to the systematic comparison of textual variants, but also to the progressive enrichment of the database and to the investigation of palaeographical, linguistic, and art-historical attribution problems. In particular, the Dante Critical Texts corpus provides the basis for developing probabilistic models of emendatio. The project aims to develop a multilayered cognitive environment in which textual philology and the philology of witnesses can be integrated, enabling a joint analysis of the material, linguistic, and iconographic dimensions of the tradition. The infrastructure can be queried through textual, codicological, linguistic, and iconographic parameters and is conceived as an open and scalable system for the study of complex textual traditions.