Faculty Profile: Prof. ChengXiang Zhai — UIUC TIMAN Group
Faculty Profile: Prof. ChengXiang Zhai — UIUC TIMAN Group
Position: Donald Biggar Willett Professor in Engineering (Full) Institution: UIUC, Dept. of Computer Science (joint: Statistics, School of Information Sciences, IGB) Lab: TIMAN Group — timan.cs.illinois.edu Report date: 2026-06-12
Research Focus
Information retrieval, text mining, user simulation, LLM for IR/AGI, text analysis.
Academic Profile
- ACM Fellow, IEEE Fellow
- ACM SIGIR Gerard Salton Award (top IR lifetime achievement award)
- 20+ years at UIUC; large lab, many alumni in IR and NLP
Key Publications (2025)
| Paper | Venue | Relevance |
|---|---|---|
| IR for Artificial General Intelligence: A New Perspective | SIGIR’25 | IR theory meets AGI |
| TINK: Text Information Navigation Kit | SIGIR’25 | Text navigation/analysis tool |
| Theory and Toolkits for User Simulation in Generative AI Era | SIGIR’25 | User simulation for LLM-based systems |
| PlugMem: plug-and-play memory for LLM agents (w/ Jiawei Han) | 2025 | LLM agent memory module |
Notable Student
- Ke Yang (PhD student): works on AI agents, language models, IR, and multimodality foundation models — suggests some lab broadening toward agents
Fit with Weijia Zhang
| Dimension | Assessment |
|---|---|
| Information retrieval | ✅ Core (but not Weijia’s focus) |
| LLM + IR / search | ⚠️ Adjacent |
| LLM agent memory | ⚠️ PlugMem is adjacent |
| Multimodal agents | ⚠️ One student doing this, not PI focus |
| Post-training (SFT / RL) | ❌ Not his area |
| GUI / VLM agents | ❌ Not his area |
Verdict
Core direction (IR, text mining) is peripheral to Weijia’s interests. Student Ke Yang’s multimodal agent work hints at some broadening, but the PI’s agenda remains IR-centric. Lower priority for Weijia compared to Ji, Hakkani-Tür, Lingming Zhang.