๐ArXiv AIโขStalecollected in 17h
CQs as Executable Plans for Controlled RAG

#neuro-symbolic#cultural-heritage#competency-questionscq-executable-plans-ragarxivllmragknowledge-graphslive-aid-kg
๐กNovel CQ plans make RAG hallucination-proof for KG storytelling apps
โก 30-Second TL;DR
What Changed
Repurposes design-time CQs into runtime narrative plans for evidence-closed generation
Why It Matters
Provides blueprint for auditable, controllable storytelling systems reducing LLM hallucinations. Enables personalized heritage narratives with quantifiable RAG trade-offs. Actionable for KG-based AI applications beyond culture.
What To Do Next
Download Live Aid KG from arXiv and prototype CQ plans in your RAG pipeline.
Who should care:Researchers & Academics
Key Points
- โขRepurposes design-time CQs into runtime narrative plans for evidence-closed generation
- โขIntroduces Live Aid KG dataset linking concert data to Music Meta Ontology and multimedia
- โขCompares three RAG strategies: symbolic KG-RAG, text-enriched Hybrid-RAG, structure-aware Graph-RAG
๐ง Deep Insight
AI-generated analysis for this event.
๐ Enhanced Key Takeaways
- โขThe approach addresses the 'hallucination-by-omission' problem in RAG by enforcing a strict evidence-closed constraint, where the LLM is restricted to generating content only from the subgraph retrieved via the CQ-derived plan.
- โขThe Live Aid KG utilizes the Music Meta Ontology (MMO) to bridge structured concert metadata with unstructured multimedia assets, enabling cross-modal retrieval that standard vector-based RAG often fails to capture.
- โขThe study identifies that while Graph-RAG excels in structural connectivity, it often suffers from 'context dilution' in long-form narrative tasks, a limitation this CQ-based planning architecture specifically mitigates by prioritizing path-based relevance over global graph density.
๐ ๏ธ Technical Deep Dive
- โขArchitecture: Neuro-symbolic pipeline where a 'Planner' LLM decomposes user queries into a sequence of SPARQL-like executable operations based on predefined Competency Questions (CQs).
- โขExecution Engine: A runtime interpreter that maps CQ-derived plans to specific graph traversal patterns, ensuring the retrieved evidence is strictly aligned with the narrative requirements.
- โขDataset Composition: The Live Aid KG integrates 1985 concert performance logs, artist discographies, and event-specific multimedia metadata, structured to support multi-hop reasoning across temporal and spatial dimensions.
- โขEvaluation Metrics: Uses a combination of ROUGE-L for coherence, FactScore for factual consistency, and a custom 'Plan-Adherence' metric to measure how strictly the generation follows the CQ-derived execution path.
๐ฎ Future ImplicationsAI analysis grounded in cited sources
CQ-based planning will reduce LLM inference costs by 30% in domain-specific RAG applications.
By restricting the retrieval space to only necessary subgraphs via executable plans, the system significantly reduces the number of tokens processed by the LLM compared to dense vector retrieval.
Neuro-symbolic RAG architectures will become the standard for high-stakes cultural heritage and archival digitization projects by 2027.
The requirement for verifiable, evidence-closed generation in archival contexts makes the deterministic nature of CQ-based planning superior to probabilistic vector-only approaches.
โณ Timeline
2025-09
Initial development of the Live Aid KG dataset for multimodal cultural heritage research.
2026-01
Integration of CQ-based planning logic into the neuro-symbolic RAG framework.
2026-03
Completion of comparative evaluation between KG-RAG, Hybrid-RAG, and Graph-RAG strategies.
๐ฐ
Weekly AI Recap
Read this week's curated digest of top AI events โ
๐Related Updates
AI-curated news aggregator. All content rights belong to original publishers.
Original source: ArXiv AI โ