HERITRACE is a web-based semantic data editor. It allows users without technical background to edit and enrich RDF metadata, while keeping track of every change with full provenance.
The full documentation is at opencitations.github.io/heritrace.
You need Docker and Docker Compose.
mkdir heritrace && cd heritrace && \
curl -o docker-compose.yml https://raw.githubusercontent.com/opencitations/heritrace/main/docker-compose.yml && \
docker compose upThe compose file starts the web application and two Virtuoso databases filled with sample data from OpenCitations Meta, ready for experimenting. On the first start, it copies an example configuration that describes this data into the config folder. Once the databases pass their health checks, the application starts at http://localhost:8000.
By default HERITRACE runs in demo mode (FLASK_ENV=demo), so you can try it without setting up ORCID authentication: every visitor is signed in as a demo user with administrator rights.
Edit docker-compose.yml: remove the dataset-db, provenance-db, config-seed, and networks blocks, remove depends_on and networks from the web service, then set DATASET_DB_URL and PROVENANCE_DB_URL to your database endpoints. On the first start, HERITRACE runs in setup mode: an administrator enters the settings, then defines the resource types in the configuration editor.
docker compose updocker compose downFor production setup with ORCID authentication, see Application settings.
- Provenance and change tracking -- every edit records who changed what and when, with full version history
- Time machine -- browse and restore previous versions of any resource
- SHACL validation -- forms and constraints are generated from SHACL shapes, with real-time validation and disambiguation
- Configuration editor -- administrators edit the SHACL shapes and the settings in the app, and HERITRACE checks them before saving
- Merging and recycle bin -- merge duplicate resources and recover deleted ones
- ORCID authentication -- restricts editing to authorized users
- Works with any RDF dataset -- connects to any SPARQL-compatible triplestore, no data migration needed
See the user guide for more details.
HERITRACE is data model agnostic. You define your domain through SHACL shapes, which drive forms, validation and presentation:
Massari, A., Peroni, S. (2027). HERITRACE: A Domain-Agnostic Framework for SHACL-Driven RDF Curation with Provenance and Change Tracking. In: Koubarakis, M., et al. The Semantic Web – ISWC 2026. ISWC 2026. Lecture Notes in Computer Science, vol 17127. Springer, Cham. https://doi.org/10.1007/978-3-032-42026-8_30
Massari, A., & Peroni, S. (2025). HERITRACE: A User-Friendly Semantic Data Editor with Change Tracking and Provenance Management for Cultural Heritage Institutions. Umanistica Digitale, 9(20), 317--340. https://doi.org/10.6092/issn.2532-8816/21218
To cite the latest version of this software (4.0.0), use this BibTeX entry:
@software{HERITRACE-4.0.0,
author = {Massari, Arcangelo},
title = {HERITRACE},
url = {https://archive.softwareheritage.org/swh:1:rev:b617ba1bef8278c27696c8212b3456dce50fda7d},
version = {4.0.0},
year = {2026}
}HERITRACE was developed as part of the PhD thesis of Arcangelo Massari, a joint doctorate between the University of Bologna (PhD in Cultural Heritage in the Digital Ecosystem, Cycle 38) and KU Leuven (Arenberg Doctoral School, Faculty of Engineering Technology), supervised by Silvio Peroni and Anastasia Dimou.
ISC. See the LICENSE file.