Skip to content

Releases: acrion/zelph

Zelph 0.5.0 Alpha Pre-Release

05 Mar 16:04
@h-b h-b
Compare
Choose a tag to compare
Pre-release

I'm excited to share the first public pre-release of zelph - a sophisticated semantic network system that allows inference rules to be defined within the network itself.

What is zelph?

zelph is a powerful foundation for knowledge representation and automated reasoning, with a special focus on efficient memory usage and logical inference capabilities. It provides a unique approach where relations are treated as first-class nodes themselves, enabling meta-reasoning about relations.

Key features in this alpha release:

  • Complete Wikidata import capability (reducing 1.4 TB to just 9.2 GB in memory)
  • Built-in contradiction detection and resolution
  • Inference engine with rules encoded within the same semantic network as facts
  • Memory-efficient data structures optimized at bit level
  • Support for multi-language node naming
  • A flexible scripting language for knowledge definition and querying

Technical Components

The zelph ecosystem includes:

  • A core C++ library providing both C++ and C interfaces
  • A single command-line binary offering both interactive usage (CLI) and batch processing
  • API functions beyond what's available in the command-line interface
  • Integration options for languages like Go and Lua through the C interface

Wikidata Integration

zelph's architecture creates a perfect alignment with Wikidata's data model:

  1. It naturally represents Wikidata's property hierarchy
  2. It can reason about properties themselves
  3. It enforces property constraints (symmetry, transitivity, inverse relationships)

zelph has processed Wikidata's vast knowledge graph and generated a tree of 4,580 pages representing entities and properties where meaningful deductions or contradictions were found. This represents the practical application of zelph's semantic inference capabilities to real-world data.

Status

This is an alpha pre-release meant for early adopters and potential contributors. While the core functionality is stable, expect evolution in APIs as we move toward a stable release.

Getting Involved