Demo Data from EDBT 2026 / reviewers in the wild / expert
Xueze Kang
dblp:376/9654
· DBLP profile ↗
1ranked-venue papers
0as first author
1since 2021 · last 2025
—ORCID · unresolved
Domains — the database's venue-derived domains; a paper can count in several
Artificial intelligence and machine learning · 1 · 1 since 2021Databases, data management, data science and information retrieval · 1 · 1 since 2021
Expertise — from the expertise taxonomy: the topics of the expert's papers under the CCF categories. A weight counts papers with recency: 1 for a paper about the topic, 0.3 when the topic is its context, halved every five years.
| Computer architecture, parallel and distributed computing, and storage systems
1 paper |
Performance modeling and evaluation · 67% Cloud and datacenter computing · 33% | |
| Databases, data mining, and information retrieval
1 paper |
Machine learning and data management · 100% |
Topics — the 3 heaviest of 4, each with the papers that count most for it
| Topic | Weight | Papers | Last | Evidence papers |
|---|---|---|---|---|
Cloud and datacenter computing › inference serving
LLM serving |
0.9 | 1 | 2025 | BurstGPT: A Real-World Workload Dataset to Optimize LLM Serving Systems · KDD (2) 2025 |
Performance modeling and evaluation
workload characterization |
0.9 | 1 | 2025 | BurstGPT: A Real-World Workload Dataset to Optimize LLM Serving Systems · KDD (2) 2025 |
Machine learning and data management › inference serving
LLM serving |
0.3 | 1 | 2025 | BurstGPT: A Real-World Workload Dataset to Optimize LLM Serving Systems · KDD (2) 2025 |
Methods — techniques the papers use, named apart from their topics
workload tracing · 1.7quality of service analysis · 1.7
| Year | Publication | Venue | Position |
|---|---|---|---|
| 2025 | BurstGPT: A Real-World Workload Dataset to Optimize LLM Serving SystemsabstractDespite efforts to improve the quality of service (QoS) and throughput in Large Language Model (LLM) serving systems, progress is often limited by the lack of publicly available real-world workloads.Consequently, evaluations usually depend on synthetic or oversimplified load patterns, and systems that appear promising in testing frequently underperform once deployed.This work presents BurstGPT, an LLM serving workload with 10.31 million traces from regional Azure OpenAI GPT services * Both authors contributed equally to this research. Yuxin Wang 0003, Yuhan Chen 0008, Xueze Kang, Yuchu Fang, Yeju Zhou, Zhenheng Tang, Xin He 0019, Qiang Wang 0022, Amelie Chi Zhou, Xiaowen Chu 0001 |
KDD (2) | 4 |