Thesis title: Annotations as Semantic Traces: A Multi-Level Ontology for Digital Hermeneutics
Annotating is considered a scholarly primitive among different fields in the humanities, and remains a central activity in the hermeneutic process. Through annotation, scholars record observations on entire works or specific passages, using theoretical concepts to construct and support interpretive arguments. In the digital domain annotations can also become a network that connects primary and secondary sources, as well as enabling scholarly discourse and interpretive debate. Nevertheless, most digital annotation systems rely on non-semantic metadata and, as a result, they are rarely integrated as first-class entities within knowledge graphs.
Using Semantic Web technologies, this study aims to define a general conceptual model for scholarly annotations based on existing and widely adopted standards in the cultural heritage and humanities domains. The model addresses the current semantic deficit in digital annotation systems, which struggle to distinguish between different conceptual levels of analysis, and enables the representation of annotations across these levels while capturing hermeneutic relationships such as scholarly agreement, disagreement, and provenance.
The results show that the model can effectively disambiguate between a target and multiple conceptual levels of the referenced entity. It also allows users to describe annotations as interpretative acts, incorporating scholarly criteria and multiple viewpoints, as well as provenance tracking, versioning, and publishing workflow management.
Two case studies validate the model across diverse humanities domains. First, the Vatican Palimpsests case study demonstrates how the model handles manuscripts containing multiple overwritten texts, distinguishing between the physical support and the distinct conceptual entities. Second, the personal manuscript archive of Charles Sanders Peirce was employed to validate the model on multimodal documents featuring visual diagrams, demonstrating its capacity to manage AI-generated annotations through scholarly editorial validation and interpretive refinement. A prototype interface plugin has been developed to enable scholars without technical expertise to create structured annotations that comply with the model. To conclude, future work will focus on extending the model to extend support to meta-annotations and intermedial relations, as well as developing semi-automatic methods for conceptual level assignment and enhanced visualisation tools for exploring annotation graphs.
keywords: annotation, Semantic Web, digital hermeneutics, ontology design, knowledge management, IIIF, semiotics