xCurator

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.

System

The xcurator system is split into two parts.

1. 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 update your data or create a new dataset for xcurator. All the code and configs are located inside the data-enrichment directory of this repository.
   
   More info: data-enrichment/README
2. Web-Application: This part of the system contains all the code and configs, which are necessary to build, run and deploy the application.
   
   More info: application/README

Tech Stack

• 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)

Credits

pixolution

In the collection exploration-feature we used the embeddings-grid technology.

Staatsbibliothek zu Berlin – Preußischer Kulturbesitz

„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/).“