Memory · Software component
Memory Deduplicator
Software componentMemoryCognition & Memoryarc:MemoryDeduplicator
A memory component that detects near-duplicate stored items by semantic similarity and merges them into a representative item with occurrence and outcome statistics.
Responsibility. Merges near-duplicate memory entries into consolidated representatives.
Also known as: Episode deduplication, Near-duplicate chunk consolidation
Relationships
is invoked by dependency
writes dependency
Design guidance
- SHOULD merge near-duplicate knowledge chunks while preserving meaningful variations, since repeated facts waste context and bias the model toward repeated content.
Quantitative guidance
As stated by the sources; verify before use.
- Example: 1,000 routine password-reset episodes merged into one representative with '95% success rate, average 5-minute resolution' (Ch5.7).
Classification
- Patterns
- Semantic near-duplicate detectionRepresentative episode plus frequency metadata
- Quality attributes
- Performance efficiency (ISO/IEC 25010)
- Risks mitigated
- Wasted storage on repeated routine episodesRedundant retrieved chunks wasting context windowRepetition bias in generation