Module gleaning_extractor 

Gleaning-based entity extraction module Gleaning-based entity extraction with TRUE LLM inference

This module implements iterative gleaning refinement using actual LLM calls, not pattern matching. Based on Microsoft GraphRAG and LightRAG research.

Expected performance: 15-30 seconds per chunk per round. For a 1000-page book with 4 gleaning rounds, expect 2-4 hours of processing time.

Structs

ExtractionCompletionStatus

Status of entity extraction completion

GleaningConfig

Configuration for gleaning-based entity extraction

GleaningEntityExtractor

Entity extractor with iterative gleaning refinement using TRUE LLM calls

GleaningStatistics

Statistics for gleaning extraction process