ZEYAN LIANG
@SoreLemon
I am a PhD student @TerrierTeam, University of Glasgow.
🎉 Glad to share that our paper "KiRAG: Knowledge-Driven Iterative Retriever for Enhancing Retrieval-Augmented Generation" (w/ @mengzaiqiao and @craig_macdonald) has been accepted at #ACL2025 as a main conference paper!
Just uploaded our paper KiRAG to Arxiv! 🚀 KiRAG integrates reasoning into retrieval, dynamically identifying and retrieving knowledge to bridge information gaps. We leverage KGs to enhance the retrieval process of iRAG. 📄:arxiv.org/abs/2502.18397 💻: github.com/jyfang6/kirag
A large contingent of our staff & students @GlasgowCS will soon be traveling to #ECIR2025. We'll be presenting our latest work in #search & #RecSys throughout the 5-day conference. Looking forward to connecting with colleagues and friends in Lucca! @ecir2025 @IDAglasgow
🎉 Our paper “Are Generative AI Agents Effective Personalized Financial Advisors?” (w/ @JavierSanzCruza @richardm_ and @iadh) has been accepted as a full paper at #SIGIR2025! Exciting times! 🇮🇹 Huge thanks to everyone who supported this work! 🙌
Great news from #SIGIR2025 — our (/w @craig_macdonald and @ntonellotto) full paper “Efficient Recommendation with Millions of Items by Dynamic Pruning of Sub-Item Embeddings” has been accepted. Nice timing too: all co-authors are now at ECIR, so we get to celebrate together.
Now accepted at #SIGIR2025! looking forward to discussing evaluation with LLMs at #ECIR2025 this week and of course in Padua! In the mean time skim this thread.
🚨 New Pre-Print! 🚨 Reviewer 2 has once again asked for DL’19, what can you say in rebuttal? We have re-annotated DL’19 in the form of classic evaluation stability studies. Work done with @maik_froebe, @hscells, @fschlatt1, @guglielm0f, @saber_zerhoudi, @macavaney, @EYangTW 🧵
Good news from #SIGIR2025: our work "Lost in Transliteration: Bridging the Script Gap in Neural IR” with @iadh and @macavaney has been accepted at the Short Paper track.
Happy to share that our (w/@graham_mcdonald and @iadh ) work "Document Similarity Enhanced IPS Estimation for Unbiased Learning to Rank" has been accepted to #SIGIR2025 as a short paper! See you in Italy, Padova. @ir_glasgow
Happy to share that our (w/ @graham_mcdonald and @craig_macdonald) work "Measuring Hypothesis Testing Errors in the Evaluation of Retrieval Systems" has been accepted to #SIGIR2025 as a short paper! 🎉Exciting times! 🇮🇹 @ir_glasgow
🚨 #ECIR2025 Demo + New Package Alert 🚨 Simplifying entry into Mechanistic Interpretability in IR /w @catherineschen @CarstenEickhoff @macavaney /🧵 soon!
MechIR: A Mechanistic Interpretability Framework for Information Retrieval Introduces a framework for diagnosing and interpreting neural IR models through mechanistic analysis. 📝arxiv.org/abs/2501.10165 👨🏽💻github.com/Parry-Parry/Me…
📢 Excited to announce that Prof. Alessandro Lenci (@AlexLenci1966) from @Unipisa and Dr. Andrew Yates (@andrewyates) from the University of Amsterdam will be delivering keynotes at the #KEIR workshop! Looking forward to hear their inspiring talks in Italy during @ecir2025. 🚀
Listening to Andrew Parry @MrParryParry give his talk entitled "In-Context Learning" or: How I learned to stop worrying and love "Applied Information Retrieval" in the LLM + Search session here at SIGIR #sigir2024. @ir_glasgow @debforit. Very interesting stuff!
After this mornings #SIGIR2024 coffee break Ill be presenting our perspective “In-Context Learning” or How I learned to stop worrying and love “Applied Information Retrieval” in M1.1 LLM+Search @SIGIRConf Time: 10:30 Room: Congressional Ballroom Paper: dl.acm.org/doi/10.1145/36…
Listening to Xuejun Chang @austin_jinqu give his talk entitled “Neural Passage Quality Estimation for Static Pruning”in the Efficiency for Search session at SIGIR #sigir2024. @ir_glasgow
Busy morning at #sigir2024 for @ir_glasgow! Just on in the "Efficiency for Search" session is @austin_jinqu presenting his work: "Neural Passage Quality Estimation for Static Pruning". 😎
#sigir2024 is off to a great start with fellow @TerrierTeam member @MrParryParry presenting his work in the first session of the day! Definitely go check out his brilliant work 'In-Context Learning' or 'How I Learned to Stop Worrying and Love Applied Information Retrieval'! 😎
I can't wait to meet everyone in Washington D.C.!🥳 I'll be speaking in the Fairness session on Tuesday and presenting a poster on Wednesday.😎Really looking forward to connecting with all of you @SIGIRConf!
A large contingent of our staff and students at @GlasgowCS are currently heading to #sigir2024! Do check out the papers and activities we are involved in. Come say hello, and chat about what we are doing in Glasgow. We are also hiring!
Happy to see this work with @debforit and Manish Chandra getting some love from awesome industry people! I'll be presenting this work in the LLM + Search session on Monday morning! Have a read in the mean time: arxiv.org/abs/2405.01116
Wrote a blog series on adaptive in-context learning (ICL) and how it connects with @vespaengine. It is a promising direction for democratizing Machine Learning(ML). blog.vespa.ai/vespa-in-conte… Work inspired by this SIGIR perspectives paper "In-Context Learning" or: How I learned…
Today, Zihan Wang @ZihanWa54274484 from the University of Amsterdam is giving an #IRTalk talk entitled "Beyond Boundaries: Towards Generalizable Named Entity Recognition Frameworks". Details at: samoa.dcs.gla.ac.uk/events/viewtal… @GlasgowCS @ir_glasgow
Today, Ferdinand Schlatt @fschlatt1 from the Friedrich Schiller University Jena is giving an #IRTalk talk entitled "Improving Cross-Encoders through Task-Specific Attention Modifications". Details at: samoa.dcs.gla.ac.uk/events/viewtal… @GlasgowCS @ir_glasgow
Was wonderful to share my ideas of the #SIGIR2024 perspective paper to the entire Glasgow team...
Today, Debasis Ganguly @debforit from The University of Glasgow is giving an #IRTalk entitled "In-Context Learning” or: How I learned to stop worrying and love “Applied Information Retrieval". Details can be found at: samoa.dcs.gla.ac.uk/events/viewtal… @GlasgowCS @ir_glasgow