📹 Demo Video

The application consists of two fundamental building blocks:

• A Node.js Express server (backend) produces ranked results with essential information such as labels, descriptions, and subgraph bindings.
• A React app (frontend) provides a user-friendly and responsive web interface.

The application requires a running SPARQL endpoint and the files representing the summary graph generated at the first off-line phase of Fed20. The routes of the backend are very lightweight, containing the OpenAPI documentation. The routes are mapped to corresponding controllers. The latter uses a common set of services. The most prominent services are as follows:

• Query service: Acts as a single entry point for all calls to the Top-k Query generator code (query generation).
• Result service: Used for the generation of the result list and the calculation of the ranking scores.
• Enrichment service: Provides all additional information for IRIs such as labels, descriptions, and image URLs. To avoid regenerating identical data and improve performance, we utilize both standard HTTP cache headers and in-memory caches implemented as Express server middleware.

The search application consists of two components:

• backend: see search-api/README.md
• frontend: see search-client/README.md

This project is licensed under the MIT License.