Supplementary Research Materials for the PhD Thesis "Vague, Incomplete, Subjective, and Uncertain Information in Digital History"

This collection contains the research materials accompanying the PhD thesis "Vague, Incomplete, Subjective, and Uncertain Information in Digital History" (Mariani, 2026). The materials support the investigation of VISU (Vague, Incomplete, Subjective, and Uncertain) information in art provenance data and document the computational and infrastructural components developed during the project. The archived materials include: (1) Art Institute of Chicago (AIC) Provenance Dataset: Structured provenance events automatically extracted from museum records; (2) NLP Models: spaCy-based models trained on manually annotated AIC provenance texts for sentence boundary detection and span categorisation; (3) PROV-A (Provenance App): A web-based application developed during the PhD for structuring provenance information as Linked Open Data, integrating automated extraction with expert validation; (4) AIC Case Study Data in PROV-A: A curated subset of AIC provenance records processed and supervised using PROV-A. -Together, these materials document the end-to-end workflow proposed in the dissertation, from automated extraction and epistemically aware modelling to human-in-the-loop validation and Linked Open Data publication.