Sean MacAvaney

he/him · Reader (Assoc. Professor) at @GlasgowCS @TerrierTeam · working at the intersection of IR&NLP · PhD from @Georgetown IRLab

[Sean MacAvaney @macavaney]

I'm delighted to have received the SIGIR Early Career Researcher Award!

Thanks to all my wonderful students, colleagues, and collaborators for their support and countless discussions about wild new ideas for the field.

https://t.co/jzG2PFMc89

[Glasgow IR Group @ir_glasgow]

Huge congratulations to @macavaney on receiving the prestigious ACM SIGIR Early Career Researcher Award in the research category! This well-deserved recognition highlights the excellence & impact of his work in the IR community 👏🎉#sigir2025
Cc @GlasgowCS @UofGlasgow @ACMSIGIR

[RAGTIME Track 🔎 TREC 2026 @trec_ragtime]

🚨 Every major AI lab is racing to build better "deep research" agents — systems that search, synthesize, and report across the web.

But how do we actually *benchmark* them?

Introducing 🧵 TREC RAGTIME — the shared task for rigorous RAG evaluation.

https://t.co/Yt9zOALedZ

[Chuan Meng @ChuanMg]

Delighted to share that our paper "Revisiting Text Ranking in Deep Research" has been accepted at #SIGIR2026, with @l1tu_0u, @macavaney, and @JeffD

We find traditional IR methods remain highly competitive in deep research

📄: https://t.co/5pxEbJdxKD
💻: https://t.co/hgZGtMgwZ4

[Craig Macdonald @craig_macdonald]

Delighted that our paper “PLAID-PRF — Pseudo-Relevance Feedback with Centroid-like Tokens in PLAID” has been accepted to #sigir2026, w/ Xiao Wang and @macavaney

[CLPsych @CLPsychOrg]

The call for papers for CLPsych 2026 collocated with ACL 2026 is out and we have a shared task that is accepting applications!
We would love to learn more about your amazing work at the intersection of NLP and clinical psychology.
https://t.co/jpoWFhyGqJ

[Craig Macdonald @craig_macdonald]

PyTerrier 1.0 has been released:
- ➡️ search pipeline validation, visualisation and verification
- 📘more documentation
- ☕️ java now optional

🆕📰https://t.co/mvxDhsmNNU
Try pip install pyterrier[all]
w/ @macavaney

[Sean MacAvaney @macavaney]

Indeed, awesome to see another method leverage rerankers with document proximity to overcome first-stage recall limitations 🤓

[Mandeep Rathee @rathee_mandeep]

Happy to see another research group, @haike_xu working in the same direction and our SlideGAR in the BRIGHT world. However, Reranker-Guided Search is not new. There are papers like Quam (WSDM'25), ORE (SIGIR'25), ReFIT(SIGIR'25), TOUR (ACL'23) that use the ranker's guidance.

[Mandeep Rathee @rathee_mandeep]

I will present our paper “Breaking the lens of the Telescope: Online Relevance Estimation over Large Retrieval Sets” at #SIGIR2025
🕰️ 10:30 AM (16.07.2025)
📍Location: GIOTTO (Floor 0)
Full Paper: https://t.co/tj1GsJBOyT
Slides: https://t.co/fpVMIaDuar

[Sean MacAvaney @macavaney]

@koko_lamba @craig_macdonald Looks like the ConstBERT code is available in the huggingface repo! Sorry that’s not clear from the Github repo.

https://t.co/P9NwoZdPTC

[Sean MacAvaney @macavaney]

@koko_lamba @craig_macdonald I’ll have to check on the ConstBERT code, sorry about that!

[Nicola Tonellotto @ntonellotto]

@joelmmackenzie kicking off the #SIGIR2025 tutorial on efficient in-memory inverted indexes w/ @macavaney and @antonio_mallia

[Andreas Chari @andreas_chari]

Starting #SIGIR2025 with @macavaney and a tutorial on "Efficient In-Memory Inverted Indexes: Theory and Practice" @ir_glasgow

[Craig Macdonald @craig_macdonald]

.@macavaney introducing the Learned Sparse tutorial at #sigir2024

[Andrew Parry @MrParryParry]

🚨 New Pre-Print! You've just added your 600th model to your negative mining pool and filtered all false negatives. Does any of this even matter when we can apply distillation?

In this work with @debforit and @macavaney, we explore data selection in modern ranking. 🧵 Below

[Sean MacAvaney @macavaney]

@beirmug @din0s_ Awesome, thanks!

[Sean MacAvaney @macavaney]

@beirmug @din0s_ Nice work! As you point out in the paper, the approach is related to existing distillation methods. I see some analysis of this in Section 6, but I would love to see a direct head-to-head comparison. Have you tried this? Might make good appendix material.

[Sean MacAvaney @macavaney]

@JackMcK1999 @orionweller @jhuclsp @GlasgowCS @ir_glasgow Awesome talk @orionweller!

[Sean MacAvaney @macavaney]

As @hscells et al say: ♻️ Reduce, Reuse, Recycle!

It's never been easier to share indexes (Terrier, Anserini, Pisa, Dense, etc.) using HuggingFace, Zenodo, etc. 🤓

[Sumit @_reachsumit]

Artifact Sharing for Information Retrieval Research

@macavaney introduces a flexible way to share artifacts like indices and models for Information Retrieval research, improving both accessibility and usability.

📝https://t.co/5XGrzOoluX
👨🏽‍💻https://t.co/w4n9Xi7BMB

[Sean MacAvaney @macavaney]

@soldni @lucy3_li @WisconsinCS @uwcdis @uwnlp @uwcse They come standard issue upon arrival 🧀

And hey, don’t look so annoyed—you love the cheesehead!

Congrats @lucy3_li!