kurra

kurra is an open source toolkit for the management of Knowledge Graph data that KurrawongAI supports commercially. It provides a range of functions to validate, query, convert, export and load data in the form of RDF and aims to assist with many of the common tasks that RDF users encounter.

It is built on RDFLib which is also an open source, low-level, RDF manipulation tool maintained by an international consortium of software experts and which KurrawongAI contributes to.

kurra has:

Detailed documentation - start here
A Python package - for use within Python scripts
A code repository - for development of kurra itself

Case Study: kurra for GA

This is an instance of kurra in use.

The Geoscience Australia manages many lists of controlled terms as Knowledge Graph vocabularies which are stored in files in RDF in a GitHub version control repository. They reload these vocabularies after changes are made to them using kurra.

kurra is used to both validate vocabularies after changes and as part of a workflow that loads them into the Fuseki RDF database that supports public access to their content, online at http://vocabs.ga.gov.au.

In addition to automated workflow tasks, GA staff can use kurra to perform many small ad hoc RDF-related tasks. They can completely remove voabularies that have conversion or loading issues, convert vocabularies in different RDF format and even view vocab concept heirarcy before loading it into the DB - see the kurra > file > hierarchy function.

Without kurra, GA staff would have to either undertake many steps for each small RDF-related tas, or re-prompt Claude or ChatGPT etc. for the same work over and over again, wasting tokens.

Our Services for kurra

training in its use
use of kurra in custom applications
- for example, GitHub Actions to load RDF data into DBs
extension of kurra
- to cover more tasks