Skip to content

Graph Concepts

DRAFT: Work in progress

This material is a work in progress, at "rough draft" stage.

Class Abstractions

The primary abstractions used in kglab are based on a small set of Python classes. These class definitions can be subclassed and extended to handle specific use cases as needed.

Knowledge Graph

kglab.KnowledgeGraph

the representation of RDF graphs, including use of semantic technologies / ontology / controlled vocabularies, and also handling property graph features

  • namespace management
  • graph construction
  • serialization
  • querying
  • validation
  • inference

Subgraph

kglab.Subgraph

  • transforms to matrix/tensor
  • label encoding
  • visualization
  • graph algorithms
  • probabilistic graph inference
  • embedding

Measure

kglab.Measure

  • descriptive statistics
  • topological analysis
  • constructing shapes
  • estimates of graph size and complexity
  • estimated computation costs

Last update: 2021-01-16