Frequently Asked Questions

This app collects data for crowdsourcing SPARQL queries for knowledge graphs. It allows users to contribute natural language questions paired with their corresponding SPARQL queries, helping to build better natural language interfaces for knowledge graphs in the social sciences and humanities domain.

Quagga is an acronym for "question answering over graphs". But Quagga ↗ is also the name of an extinct subspecies of the plains zebra. A bit like Quaggas, benchmark datasets for knowledge graphs question answering in the social sciences and humanities domain are also rare and hard to find. That's why we have set out to build a comprehensive benchmark da, and called it Quagga.

The platform supports knowledge graphs in the social sciences and humanities domain. You can browse existing submissions to see which endpoints are currently being used, or you can add new endpoints by providing the SPARQL endpoint URL along with a name and description for the knowledge graph. If a live SPARQL endpoint is not available, you will be asked to provide a link to a knowledge graph dump.

To contribute, you need to log in using GitHub, ORCID or OPERAS ID authentication. Once logged in, you can access the Contribute page where you can submit natural language questions along with their corresponding SPARQL queries for various knowledge graph endpoints. You can also submit questions without SPARQL queries, if you don't know how to formulate your question as a SPARQL query.

Yes, you can modify your own submissions. When you're logged in and are viewing your submissions, you'll see a "Edit Query" button next to each of your entries. This allows you to update the SPARQL query as needed.

Yes, we are planning to publish the collected data as an open source benchmark dataset, once we reach a sufficient data volume. Also, we will acknowledge all contributors in the dataset documentation.