Sema: A High-performance System for LLM-based Semantic Query Processing
Summary: SemaSQL makes LLM operators first-class in DuckDB, embedding natural-language expressions in SQL for end-to-end optimization. Constraint-aware rewriting and adaptive execution reorder/fuse operators and batch prompts, yielding 2–10× speedups under accuracy and token/latency objectives. (summarized by gpt-5.6-luna on Aug 28 2026)
Incoming Non-self Citations Over Time
No non-self incoming citations found for this paper in this database.
Authors
- 1. Kangkang Qi (Beijing Institute of Technology)
- 2. Dongyang Xie (Wuhan University)
- 3. Wenbo Li (Wuhan University)
- 4. Hao Zhang (Chinese University of Hong Kong)
- 5. Yuanyuan Zhu (Wuhan University)
- 6. Jeffrey Xu Yu (Hong Kong University of Science and Technology)
- 7. Kangfei Zhao (Beijing Institute of Technology)
BibTeX Citation
@article{qi_vldb26,
title = {{Sema: A High-performance System for LLM-based Semantic Query Processing}},
author = {Qi, Kangkang and Xie, Dongyang and Li, Wenbo and Zhang, Hao and Zhu, Yuanyuan and Yu, Jeffrey Xu and Zhao, Kangfei},
journal = {PVLDB},
series = {{VLDB} '26},
volume = {19},
number = {11},
pages = {3231--3244},
doi = {10.14778/3836663.3836685},
url = {https://doi.org/10.14778/3836663.3836685},
year = {2026}
}Incoming Citations (Sorted by Pagerank)
Showing 0 of 0 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 28 of 28 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
Previous
Page 1 / 1
Next