Skyline Retrieval meets Set-Cover Chunk Merging: A Cost-Effective RAG-Sketch for Long-Context LLM QA

Summary: RAG-Sketch combines schema-driven query decomposition with skyline retrieval to select Pareto-optimal chunks balancing similarity and key-information coverage. A greedy set-cover merger removes redundant context access, improving long-context QA accuracy 24.5% while cutting token cost 58.5%. (summarized by gpt-5.6-luna on Jul 26 2026)

Paper ID: 7484
Venue: SIGMOD
Year: 2026
Pagerank: 5.093636e-05
Overall Rank: 10,290 | 29.41%
DOI: 10.1145/3802111

Incoming Non-self Citations Over Time

No non-self incoming citations found for this paper in this database.

Authors

1. Xinyi Zhu (Hong Kong University of Science and Technology)
2. Haoyang Li (Hong Kong Polytechnic University)
3. Yongqi Zhang (Hong Kong University of Science and Technology)
4. Lei Chen (Hong Kong University of Science and Technology)

BibTeX Citation

@inproceedings{zhu_sigmod26,
        title = {{Skyline Retrieval meets Set-Cover Chunk Merging: A Cost-Effective RAG-Sketch for Long-Context LLM QA}},
        author = {Zhu, Xinyi and Li, Haoyang and Zhang, Yongqi and Chen, Lei},
        series = {{SIGMOD} '26},
        booktitle = {Proceedings of the {ACM} {SIGMOD} International Conference on Management of Data},
        publisher = {Association for Computing Machinery},
        doi = {10.1145/3802111},
        url = {https://dl.acm.org/doi/10.1145/3802111},
        year = {2026}
}

Incoming Citations (Sorted by Pagerank)

Showing 0 of 0 citing papers.

Rank	Citing Paper	Year	Venue	Pagerank

Outgoing Citations (Sorted by Pagerank)

Showing 11 of 11 cited papers.

Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.

Rank	Cited Paper	Year	Venue	Pagerank
713	Language Models Enable Simple Systems for Generating Structured Views of Heterogeneous Data Lakes	2024	VLDB	0.00014672521
2,816	Cache-Craft: Managing Chunk-Caches for Efficient Retrieval-Augmented Generation	2025	SIGMOD	8.0959781e-05
4,045	Logical and Physical Optimizations for SQL Query Execution over Large Language Models	2025	SIGMOD	6.9394654e-05
4,886	Chat2Data: An Interactive Data Analysis System with RAG, Vector Databases and LLMs	2024	VLDB	6.4614955e-05
6,101	LLM for Data Management	2024	VLDB	5.9774273e-05
6,254	In-depth Analysis of Graph-based RAG in a Unified Framework	2025	VLDB	5.9405174e-05
7,826	AlayaDB: The Data Foundation for Efficient and Effective Long-context LLM Inference	2025	SIGMOD	5.53654e-05
7,871	TSGAssist: An Interactive Assistant Harnessing LLMs and RAG for Time Series Generation Recommendations and Benchmarking	2024	VLDB	5.5271792e-05
8,343	An Interactive Multi-modal Query Answering System with Retrieval-Augmented Large Language Models	2024	VLDB	5.448279e-05
9,306	SagaLLM: Context Management, Validation, and Transaction Guarantees for Multi-Agent LLM Planning	2025	VLDB	5.289545e-05
9,310	LLM-PBE: Assessing Data Privacy in Large Language Models	2024	VLDB	5.289545e-05

Semantically Similar Papers

#	Overall Rank	Paper	Year	Venue
1	10,382	LLM-Powered Interactive Graph Search: A Scalable and Practical Approach	2026	SIGMOD
2	6,765	AquaPipe: A Quality-Aware Pipeline for Knowledge Retrieval and Large Language Models	2025	SIGMOD
3	8,343	An Interactive Multi-modal Query Answering System with Retrieval-Augmented Large Language Models	2024	VLDB
4	10,437	Chatty-KG: A Multi-Agent AI System for On-Demand Conversational Question Answering over Knowledge Graphs	2026	SIGMOD
5	10,405	AixelAsk: A Stepwise-Guided Retrieval and Reasoning Framework for Large Table QA	2026	SIGMOD
6	10,499	SchemaRAG: A Schema-aware Retrieval-Augmented Generation Framework for Text-to-SQL	2026	SIGMOD
7	10,663	Dialogue Benchmark Generation from Knowledge Graphs with Cost-Effective Retrieval-Augmented LLMs	2025	SIGMOD
8	2,816	Cache-Craft: Managing Chunk-Caches for Efficient Retrieval-Augmented Generation	2025	SIGMOD
9	10,459	From Prefix Cache to Fusion RAG Cache: Accelerating LLM Inference in Retrieval-Augmented Generation	2026	SIGMOD
10	6,254	In-depth Analysis of Graph-based RAG in a Unified Framework	2025	VLDB