The web app xCurator is a tool that enriches object datasets from museum databases, making them searchable and contextualized. With a variety of AI technologies and an intuitive interface, it empowers users to explore museum collections and recognize connections. The integrated editor allows users to assemble objects into stories.
The xcurator system is split into two parts.
- Data Enrichment: This part of the system is responsible for importing, cleaning and enriching the data.
You need to run it if you want to
updateyour data orcreatea new dataset for xcurator. All the code and configs are located inside the data-enrichment directory of this repository.
More info: data-enrichment/README - Web-Application: This part of the system contains all the code and configs, which are necessary to
build,runanddeploythe application.
More info: application/README
- Frontend: TypeScript | Next.js | React | GraphQL (Apollo) | yarn
- Backend: Java (OpenJDK) | Spring Boot | GraphQL (DGS Netflix) | Gradle
- AI | Data: Python | Pandas | PyTorch | Poetry
- Infrastructure: MongoDB | Elasticsearch | Keycloak | Docker (Compose)
In the collection exploration-feature we used the embeddings-grid technology.
„The Named Entity Recognition and Entity Linking system used was developed by the Staatsbibliothek zu Berlin – Preußischer Kulturbesitz (https://staatsbibliothek-berlin.de/) in the BMBF-funded project QURATOR – Curation Technologies (https://qurator.ai/).“