KGraphPy – Knowledge Graph DataOps for Python

Table of Contents

• Overview
• Scope and Focus
• Core Capabilities
• Interoperability and DataOps Extensions
• Relationship to Other Libraries
• Licensing
• Accreditation

Overview

KGraphPy is a Python library for working with knowledge graphs on top of RDFLib.

It provides DataOps-oriented extensions for standards-based graph data, with a particular focus on CIM / CGMES–style model exchange, change management, and dataset packaging. KGraphPy is designed to complement RDFLib by adding practical capabilities needed in real-world data and model pipelines, while remaining fully compatible with RDFLib’s APIs and ecosystem.

KGraphPy is intended to serve as a foundational building block for higher-level DataOps and ModelOps workflows, including validation, verification, analytics, and human-in-the-loop correction.

Scope and Focus

KGraphPy focuses on:

• Working with RDF-based knowledge graphs in Python

• Supporting standards-based data exchange, with CIM / CGMES as the primary initial focus

• Providing interchange, packaging, and change-management support commonly required in operational environments

• Integrating cleanly with schema, validation, and analytics tooling

While CIM / CGMES is the primary driver, KGraphPy is designed to be reusable across other standards and profiles, including IEC 61850, BIM/IFC, GeoSPARQL-based datasets, EU SEMIC profiles, and Industrial Data Ontology (IDO).

Core Capabilities

KGraphPy builds directly on RDFLib and provides:

• In-memory knowledge graph handling (and supported backends via RDFLib)

• Parsing and serialisation of RDF in multiple syntaxes (RDF/XML, Turtle, TriG, JSON-LD, etc.)

• SPARQL 1.1 querying over local graphs and remote endpoints via RDFLib stores

• Seamless integration with schema and validation tooling such as pySHACL and LinkML

Interoperability and DataOps Extensions

KGraphPy adds practical DataOps-oriented extensions, including:

• CIMXML support Handling of CIM RDF/XML conventions used in IEC CIM / CGMES exchanges

• CIMXML header support Explicit handling of metadata and header graphs as defined in CGMES packaging

• DifferenceSet support

• Creation, parsing, and application of CIM DifferenceSets

• Comparison of model versions (base vs. changed state)

• Incremental exchange and change-oriented DataOps workflows

• Dataset-level metadata support Support for modern DCAT-3–style dataset headers, enabling explicit dataset metadata graphs

• Compressed artefact handling Read/write support for compressed packages (e.g. ZIP), including common profile instance set bundle layouts

• Future-ready packaging Designed to accommodate evolving CIM exchange formats, including JSON-LD and packaging

Relationship to Other Libraries

KGraphPy is intentionally lightweight and composable:

• RDFLib provides the core RDF graph engine and SPARQL capabilities

• LinkML can be used to define schemas, profiles, and generate typed Python artefacts

KGraphPy does not replace RDFLib; it extends it with operational and standards-oriented capabilities.

Licensing

KGraphPy is released under the Apache License, Version 2.0.

This allows free use, modification, and redistribution in open-source, research, and commercial contexts, subject to the terms of the license.

Accreditation

KGraphPy is built on top of RDFLib, and we explicitly acknowledge the RDFLib project and its contributors for providing the foundational RDF graph capabilities that make this library possible.

• RDFLib project: https://github.com/RDFLib/rdflib

In addition, we acknowledge and accredit:

• UCA CIMug

• The teams responsible for the development of Common Information Model (CIM) for Electricity, which define the information models, application profile and exchange format that are suppored by KGraphPy’s interoperability features.

KGraphPy builds on this work but remains an independent, open-source Python library.