Class DocumentQualitySummary
- Namespace
- LMKit.Retrieval.ChunkQuality
- Assembly
- LM-Kit.NET.dll
Aggregated gating totals for one document, produced by Complete(). Useful for ingestion telemetry and for detecting documents whose text extraction produced garbage wholesale.
public sealed class DocumentQualitySummary- Inheritance
-
DocumentQualitySummary
- Inherited Members
Properties
- ChunksDropped
Chunks rejected as semantically worthless.
- ChunksEvaluated
Total source chunks submitted to the session.
- ChunksFlagged
Chunks kept out of the index with conflicting evidence, for review.
- ChunksIndexed
Chunks that became (or joined) an indexed unit.
- ChunksMerged
Chunks folded into a neighboring unit by the merge policy.
- DominantIssues
The union of the issues observed on dropped and flagged chunks.
- LooksExtractionFailed
true when the document's chunks were overwhelmingly dropped as extraction noise (see ExtractionFailedDropRatio), which usually means the source document's text layer is unusable and the document should be re-extracted (for example with OCR) rather than indexed.