KGraphLang

KgraphLang is a knowledge graph query language designed for LLMs.

The repo is: https://github.com/vital-ai/kgraphlang

Further documentation: https://vital-ai.gitbook.io/kgraphlang

KGraphLang can be used with Ensemble Reasoning to enable reasoning to generate knowledge graph queries and process results during inference.

The kgraphlang query can be used with KGraphService to process the query, or another implementation can be used. For KGraphService, the predicates available are mapped to KGraphService queries, which may include local caching to improve performance.

The implementation of a predicate may handle tabling or caching of data internally, or a query re-writing / optimization process may replace a predicate or set of predicates with an optimized replacement via processing the query parse AST.