| --- |
| title: README |
| emoji: 🔍 |
| colorFrom: pink |
| colorTo: blue |
| sdk: static |
| pinned: false |
| --- |
| |
| # THUIR — Information Retrieval Lab, Tsinghua University |
|
|
| 清华大学信息检索课题组 |
|
|
| THUIR is the Information Retrieval research group at Tsinghua University, founded in the 1990s by Prof. Shaoping Ma as one of China's earliest IR research groups. We work on the theories, algorithms, and applications of intelligent search and recommendation systems, aiming to bridge the gap between user information needs and available information resources. |
|
|
| We are ranked **#1 globally in Web & Information Retrieval on [CSRankings](https://csrankings.org)**, with hundreds of papers published in top-tier venues (SIGIR, WWW, KDD, ACL, NeurIPS, etc.). |
|
|
| ## Team |
|
|
| - 4 Principal Investigators |
| - ~30 current graduate students |
| - 30+ Ph.D. and 50+ M.S. graduates to date |
|
|
| ## Research Areas |
|
|
| - Search & retrieval (dense/sparse retrieval, ranking, RAG) |
| - Recommendation systems |
| - User behavior modeling & evaluation |
| - LLM-based agents and memory |
|
|
| ## Open-Source Resources |
|
|
| - [ReChorus](https://github.com/THUIR/ReChorus) — a unified, flexible recommendation algorithm toolkit |
| - [ULTRA](https://github.com/ULTR-Community/ULTRA) — unbiased learning-to-rank algorithms |
| - [T2Ranking](https://huggingface.co/datasets/THUIR/T2Ranking), [Qilin](https://huggingface.co/datasets/THUIR/Qilin), [MemoryBench](https://huggingface.co/datasets/THUIR/MemoryBench) — retrieval & memory benchmarks released on the Hub |
|
|
| ## Links |
|
|
| - Website: [www.thuir.cn](http://www.thuir.cn) |
| - GitHub: [github.com/THUIR](https://github.com/THUIR) |
| - Twitter/X: [@thuir_lab](https://twitter.com/thuir_lab) |
|
|
| ## Join Us |
|
|
| We recruit ~3-5 Ph.D. students, ~1 M.S. student, and ~1 postdoc per year. See our website for openings. |
|
|
