Bridging Prosopographical Research Through Pragmatic Digital Integration – The Prosopographical Research Platform Austria (PFP)
The Prosopographical Research Platform Austria (PFP) offers a novel, non-invasive approach to integrating heterogeneous prosopographical datasets without requiring projects to abandon their existing tools or data models. PFP addresses a core challenge in digital prosopography: how to connect dispersed, often incompatible datasets while preserving their autonomy, provenance, and scholarly value. Rather than imposing a monolithic system, PFP embraces a pragmatic, modular architecture that harmonizes data through open standards – particularly CIDOC CRM, RDF, and SHACL – while leaving source datasets untouched in their original formats.
At its heart, PFP functions as a knowledge graph that aggregates and links entities across projects. Each dataset is stored in a versioned, named graph, ensuring traceability and non-destructive updates. A provenance layer records the origin of every statement, while a consolidation layer resolves the problem of duplicate entities (e.g., the same historical person described in multiple datasets under different identifiers). Using “owl:sameAs” relations and shared authority files (such as GND identifiers), PFP generates unified UUIDs for entities, allowing the platform to expose a single, aggregated view of an entity while retaining links to the original records. For example, the writer Arthur Schnitzler appears in both the Österreichisches Biographisches Lexikon (ÖBL) and Personen der Moderne (PMB) datasets; PFP recognizes these as the same individual and merges their data into a single, queryable entity.
The platform’s technical design reflects lessons learned from prior digital humanities projects (most notably InTaVia: https://intavia.eu): (1) reuse established tools (e.g., GitLab for versioning, Prefect for ETL workflows), (2) validate rigorously (via SHACL shapes to enforce data consistency), and (3) minimize disruption to running systems. The only custom component, RDFProxy, bridges the gap between the SPARQL endpoint and user-friendly interfaces by mapping query results to structured, nested data models. This modular design allows for independent component replacement while supporting key standards: RDF for the database backend and JSON APIs for the frontend.
PFP is still heavily under development, but nonetheless already integrates data from 11 datasets created at the Austrian Academy of Sciences (OeAW). The data is accessible via GIT (https://gitlab.oeaw.ac.at/acdh-ch/pfp/pfp-source-data), SPARQL (https://qlever-ui.acdh-dev.oeaw.ac.at/pfp), and a beta version of the API (https://pfp-api.acdh-dev.oeaw.ac.at/docs). The PFP knowledge graph currently contains well over 140,000 person proxies, about 12,000 institution proxies, and nearly 60,000 place proxies.
The presentation will give a brief overview of the technical setup and focus on application in humanities research and future development.
*
Matthias Schlögl is a lead research software engineer at the ACDH. He has more than 10 years of experience in using and developing digital tools for social sciences and humanities research. He received a diploma in International Development from the University of Vienna and worked at various national and international institutions, including the University of Bath and the Berlin Social Science Center, before joining the ACDH in 2015. At the ACDH he is currently coordinating the Prosopography and Networks team within the research unit DH Research & Infrastructure.
E-mail: matthias.schloegl@oeaw.ac.at
