Unstructured Textual Data Retrieval TRIZ Cases
- Multi-Phase Query Plan Caching for Faster Database Compilation
A multi-phase cache stores reusable compiled query plans to cut repeated parsing and optimization, improving response time and resource use.
- Posted Information Verification Using Event Influence Ranges
Assesses posted information reliability by matching event-based geographic and time influence ranges against metadata and trusted sources.
- Knowledge Graph Re-Prompting for Consistent RAG Context Selection
A knowledge graph guides follow-up prompts to resolve ambiguous RAG chunk matches and improve LLM response consistency and accuracy.
- Account Consolidation Using Similarity Clustering for Access Monitoring
Similarity-based account clustering links application accounts to one entity, improving access monitoring and abnormal usage detection.