aarondb/engine/cognitive
Values
pub fn solve(
db_state: state.DbState,
concept: ast.Part,
context: ast.Part,
threshold: Float,
engram_var: String,
ctx: dict.Dict(String, fact.Value),
as_of_tx: option.Option(Int),
as_of_valid: option.Option(Int),
) -> List(dict.Dict(String, fact.Value))
Solve AaronDB’s local cognitive clause over explicit engram facts.
An entity matches only when it has active engram/concept and
engram/context facts matching the supplied parts. Its relevance is the
maximum active numeric engram/relevance value; an entity with no relevance
fact has relevance 1.0, while one with only non-numeric relevance facts has
relevance 0.0. The threshold comparison is inclusive.
This module performs no learning, decay, vector embedding, or external retrieval. Relevance changes only through ordinary fact assertions and retractions, and temporal bases select the active facts visible at that basis.