Set up a Blazegraph cloud endpoint with a loaded Wikidata dump

Wikidata, a sister project of Wikipedia, is a graph database which aims to model all real-world knowledge. It can be used via the open SPARQL endpoint, which is run by a Blazegraph backend.

The problem with this endpoint is a 1-minute timeout. We need to run heavy queries which take more time than one minute. Thus, we need to set up our own backend to run the queries.

Your tasks:

* Set up a container / VM in a cloud (like Amazon)
* Install Blazegraph on this container
* Download a Wikidata dump (probably you need it in the Turtle format, i.e. latest-all.ttl)
* Import this dump in Blazegraph
* Configure a SPARQL endpoint, similar to the one on query wikidata. It does not need to be the exact same GUI, but we need a JSON endpoint as well as a GUI endpoint to manually test the queries.
* Secure the container and Blazegraph, so that only we have access to it
* If possible, we would like to tweak the resources of the VM (CPU, memory), to find out how mcuh we can improve the performance of the SPARQL queries.

Every step must be documented, so that we can repeat it later.

The best option would be using a pre-existing service (some kind of Blazegraph-as-a-service).

In your offer, please describe the tools you would be using (which cloud provider etc.)

Please note the Budget of 5$ is not meant seriously. We dont know how much this would cost, but of course way more than 5$. If its too expensive, we wont do it, but this does not just depend on the number, but also on the offered quality.
About the recuiter
Member since Mar 14, 2020
Shubham Goel
from Zacatecas, Mexico

---

---

Looking for help? Checkout our video tutorial
How to search and apply for jobs

How to apply? Do you have more questions about the Job?
See frequently asked questions