iiinova — where your digital self keeps evolving

Reposted by Glasgow IR Group

Glasgow IR Group @irglasgow.bsky.social · 25/12/2025

🎄 PyTerrier Advent 25/25: To wrap up the our advent series, we'd like thank the contributors shown below, and to the many others who support the PyTerrier ecosystem! #WorldChangersTogether

011

Reposted by Glasgow IR Group

Glasgow IR Group @irglasgow.bsky.social · 24/12/2025

🎄 PyTerrier Advent 24/25: Removing low-quality docs can boost search quality and cut indexing costs. Our SIGIR’24 paper QT5 trains a T5 model to filter passages at indexing time—easy to integrate, and works with dense, PISA, or SPLADE indexes too.

143

Reposted by Glasgow IR Group

Glasgow IR Group @irglasgow.bsky.social · 23/12/2025

🎄PyTerrier Advent 23/25: You’ve done retrieval, but the results seem too homogeneous. Use a diversification reranker. Shown below is the implicit MMR diversification approach, instantiated on a BM25 or dense retrieval, but even an explicit approach like xQuAD (c.f. Rodrygo Santos) is easy to write.

133

Reposted by Glasgow IR Group

Glasgow IR Group @irglasgow.bsky.social · 22/12/2025

🎄 PyTerrier Advent 22/25: A more complex pipeline—knowledge-graph–enhanced RAG from our EMNLP 2024 paper TRACE. We build a KG over retrieved docs, then use a transformer to reason over triples for better QA. This pipeline instantiation uses a cache (see 20th advent) on LLM-based KG extraction.

144

Reposted by Glasgow IR Group

Glasgow IR Group @irglasgow.bsky.social · 21/12/2025

🎄PyTerrier Advent 21/25: Bounded recall blues got you down? You can use Adaptive Retrieval techniques, like GAR, LADR, and LAFF, to efficiently surface missing relevant documents.

122

Reposted by Glasgow IR Group

Glasgow IR Group @irglasgow.bsky.social · 20/12/2025

🎄PyTerrier Advent 20/25: You can think of every PyTerrier transformer as a function - mapping from one dataframe type to another. This makes them easily cachable, courtesy of pyterrier_caching. We have cache object for retrievers, rerankers, or even indexing-time transformations (e.g. Doc2Query)

122

Reposted by Glasgow IR Group

Glasgow IR Group @irglasgow.bsky.social · 19/12/2025

🎄 PyTerrier Advent 19/25: PyTerrier-RAG brings agentic RAG to your workflows with support for SOTA methods like Search-R1 and R1-Searcher, to combine retrievers and reasoning. You could even swap BM25 out for dense or LSR retriever.

101

Reposted by Glasgow IR Group

Glasgow IR Group @irglasgow.bsky.social · 18/12/2025

🎄 PyTerrier Advent 18/25: In RAG, the reader runs the LLM—but your pipeline shouldn’t depend on the LLM stack. PyTerrier-RAG separates Reader from Backend, letting you swap vLLM ↔ HF with one line while keeping the same pipeline (and even share a Backend with other stages).

101

Reposted by Glasgow IR Group

Glasgow IR Group @irglasgow.bsky.social · 17/12/2025

🎄 PyTerrier Advent 17/25: FlexIndex simplifies dense retrieval by supporting FAISS, Voyager, FlatNav & more. It auto-builds data structures, reuses vector stores to cut storage costs, and offers one familiar API for many retrievers.

101

Reposted by Glasgow IR Group

Glasgow IR Group @irglasgow.bsky.social · 16/12/2025

🎄PyTerrier Advent 16/25: Speaking of Learned Sparse Retrieval, PyTerrier has bindings to two backend search engines that provide blazing-fast retrieval over sparse vectors: PISA and BMP. You can see that we really work to keep the look-and-feel uniform between implementations

121

Reposted by Glasgow IR Group

Glasgow IR Group @irglasgow.bsky.social · 15/12/2025

🎄PyTerrier Advent 15/25: A very well-known learned sparse method is SPLADE. Our pyt_splade plugin makes it easy to use SPLADE by formulating Terrier indexing & retrieving pipelines that are composed with a SPLADE encoder, adding extra columns (e.g. query_toks). Try it 👉 github.com/cmacdonald/p...

112

Reposted by Glasgow IR Group

Glasgow IR Group @irglasgow.bsky.social · 14/12/2025

🎄 PyTerrier Advent 14/25: So we’ve seen sparse and dense retrieval in PyTerrier. Some folk recommend hybrid retrieval – e.g. reciprocal rank fusion (RRF) of sparse and dense results. We have an easy pipeline component that combine two sets of results by RRF (w/ a pretty schematic by Sean MacAvaney)

101

Reposted by Glasgow IR Group

Glasgow IR Group @irglasgow.bsky.social · 10/12/2025

🎄 PyTerrier Advent 10/25: Dense retrieval often improves with pseudo-relevance feedback (Rocchio-style). In PyTerrier_DR it’s easy: encode query, retrieve docs, a transformer to mix doc vectors w/ the query vector, and then re-retrieve. pyterrier.readthedocs.io/en/latest/ex...

101

Reposted by Glasgow IR Group

Glasgow IR Group @irglasgow.bsky.social · 13/12/2025

🎄 PyTerrier Advent 13/25: Doc2query expands docs with generated queries, but can hallucinate. Our ECIR’23 paper Doc2query-- (aka "minus minus") filters generated queries using a cross-encoder before indexing. PyTerrier pipeline: generate→score→filter→index. 📄https://arxiv.org/pdf/2301.03266

133

Reposted by Glasgow IR Group

Glasgow IR Group @irglasgow.bsky.social · 12/12/2025

🎄 PyTerrier Advent 12/25: Beyond dense retrieval, learned sparse methods like Doc2Query expand docs with predicted queries before indexing. Our pyterrier_doc2query plugin makes this easy for any corpus—perfectly intuitive as PyTerrier’s pipelines can be applied at indexing time too!

101

Reposted by Glasgow IR Group

Glasgow IR Group @irglasgow.bsky.social · 11/12/2025

🎄PyTerrier Advent 11/25: Want to use an external search services with PyTerrier? No problemo! It has integrations with APIs for Semantic Scholar, ChatNoir (thanks to Jan Heinrich Merker!), Pinecone, and others!

111

Glasgow IR Group @irglasgow.bsky.social · 10/12/2025

101

Reposted by Glasgow IR Group

Glasgow IR Group @irglasgow.bsky.social · 09/12/2025

🎄PyTerrier Advent 9/25: Yesterday—dense retrieval with E5 via PyTerrier_DR. Today—RAG! PyTerrier_RAG readers generate answers from retrieved docs. Example: a FiD reader over E5 results, w/ & w/o monoT5 reranking 👇 Check the full notebook to see impact on answer quality. github.com/terrierteam/...

001

Reposted by Glasgow IR Group

Glasgow IR Group @irglasgow.bsky.social · 08/12/2025

🎄PyTerrier Advent 8/25: Beyond sparse! PyTerrier_dr adds dense indexing & retrieval. Instantiate an encoder model, compose with FlexIndex. Retrieval is identical. Support models include: ANCE, TCT-ColBERT, BGE, E5, or any SentenceTransformer model. 👉 github.com/terrierteam/...

134

Reposted by Glasgow IR Group

Glasgow IR Group @irglasgow.bsky.social · 06/12/2025

🎄PyTerrier Advent 6/25: Lets see some more neural rerankers: MonoT5 & DuoT5 (DuoT5 is a more costly pairwise reranker). We use PyTerrier's succinct rank cutoff operator (%) to cut the nbr of retrieved docs reranked by each of the MonoT5 & DuoT5 stages. Try it out: github.com/terrierteam/...

111

Reposted by Glasgow IR Group

Glasgow IR Group @irglasgow.bsky.social · 05/12/2025

🎄PyTerrier Advent 5/25: Comparing pipelines (e.g., BM25 vs RM3)? Use pt.Experiment(baseline=…) to see MAP gains/losses per query and paired t-test p-values. Multiple-test corrections are supported too. 📄https://pyterrier.readthedocs.io/en/latest/experiments.html#significance-testing

111

Reposted by Glasgow IR Group

Javier Sanz-Cruzado @javiersanzcruza.bsky.social · 03/12/2025

Paper alert 🚨"Investors Are (Not) Always Right: A Comparison of Transaction-Based and Profitability-Based Metrics for Financial Asset Recommendations", with @richardmcc.bsky.social, Nikos Droukas, @craigmacdonald.bsky.social & @iadhounis.bsky.social has been accepted at ACM TOIS! 🧵(1/5)

143

Reposted by Glasgow IR Group

UofGNews @uofgnews.bsky.social · 02/12/2025

PyTerrier, a software platform developed at @uofgcompsci.bsky.social which helps facilitate the development of AI-powered search engines, has won a national award from @wearebcs.bsky.social! Read more here: www.gla.ac.uk/news/headlin...

Dr Sean MacAvaney receives the Best Search Project of the Year Award at the BCS Search Industry Awards on Wednesday 26 November.

094

Glasgow IR Group @irglasgow.bsky.social · 01/12/2025

🎄We want to try something new and fun this year – an “Advent Calendar” of PyTerrier pipelines 🤓 We’ll kick it off with *the* baseline: BM25 on MSMARCO. One line to download a pre-built index, one line to make a BM25 retriever, one line to search.

184

Glasgow IR Group @irglasgow.bsky.social · 28/11/2025

We are happy to sponsor the Keith Van Rijsbergen (KvR) Award at #ECIR2026 📢 𝗡𝗼𝗺𝗶𝗻𝗮𝘁𝗶𝗼𝗻𝘀 𝗮𝗿𝗲 𝗻𝗼𝘄 𝗼𝗽𝗲𝗻! Details and Submission Form at @ecir2026.eu website: lnkd.in/emecdsXm 🗓 𝗗𝗲𝗮𝗱𝗹𝗶𝗻𝗲: 𝟭𝟱 𝗝𝗮𝗻𝘂𝗮𝗿𝘆 𝟮𝟬𝟮𝟲 www.linkedin.com/posts/glasgo...

linkedin.com

Keith van Rijsbergen Award | Glasgow Information Retrieval Group

𝗞𝘃𝗥 𝗔𝘄𝗮𝗿𝗱 @ 𝗘𝗖𝗜𝗥 𝟮𝟬𝟮𝟲 We are proud to sponsor the Keith Van Rijsbergen (KvR) Award at ECIR 2026. Keith van Rijsbergen is a pioneer of modern Information Retrieval, with a long history at the Univers...

065

Reposted by Glasgow IR Group

Kate Simpson @kilmun.bsky.social · 18/11/2025

The brilliant @annarezk.bsky.social from @uofglasgow.bsky.social presenting on stakeholder research as part of the RAI funded Participatory Auditing Harm Workbenches and Methodologies project at the #RAIAHM in Manchester. Innovative and significant work #PHAWM #RAI2025 #AIAuditing

A person in a dark top with long hair stands at a lecturn in the bottom right of the image. Behind them is a large screen showing stakeholder research in AI auditing techniques

042

Glasgow IR Group @irglasgow.bsky.social · 11/07/2025

📢 We're off to #SIGIR2025 in Padova! A large contingent of our students & staff will be at the main conference + tutorials, workshops & #ICTIR2025. Let’s connect if you’re around! 🤝 Also: we're hiring in #FinTech — DM us if you're interested! #IR #Hiring #Research #ACMSIGIR

0103

Reposted by Glasgow IR Group

mengzaiqiao.bsky.social @mengzaiqiao.bsky.social · 19/05/2025

🎉 An open-source AI Agent project, EvoAgentX, led by @uofglasgow.bsky.social PhD students, has received 260+ ⭐️ on GitHub in just 5 DAYS!🔥 🤖 The project aims to build a Self-Evolving Ecosystem of AI Agents. @irglasgow.bsky.social @idaglasgow.bsky.social #AI #EvoAgentX

131

Glasgow IR Group @irglasgow.bsky.social · 10/07/2025

Huge congratulations to our brilliant PhD graduates: @tjaenich.bsky.social, @mariavlachou.bsky.social, and @apetrov.bsky.social 🎉 It’s been a joy having you @uofgcompsci.bsky.social - We’re so proud of all you’ve achieved and can’t wait to see what amazing things you’ll do next! 👏 #PhDGraduation

072

Reposted by Glasgow IR Group

grahammcdonald.bsky.social @grahammcdonald.bsky.social · 24/04/2025

Congratulations to @tjaenich.bsky.social who defended his PhD Thesis Fair Group-Based Exposure of Documents in Ranked Search Results 👏👏 Was a pleasure supervising your PhD along with @iadhounis.bsky.social. @irglasgow.bsky.social Thanks to the examiners Dr. Sole Pera and Christos Anagnostopoulos!

0126

Glasgow IR Group @irglasgow.bsky.social · 13/04/2025

Arrivederci #ecir2025 … and thanks for all the fish - see you all in Delft!

Members of the Glasgow IR Group waving goodbye at the closing of the last day of ECIR 2025

081

Glasgow IR Group @irglasgow.bsky.social · 09/04/2025

Congratulations to Manish Chandra, @gdebasis.bsky.social and @iadhounis.bsky.social for being awarded the Best Paper Award at #ECIR2025 for their work entitled “One size doesn’t fit all: Predicting the number of examples for in-context learning”. The paper will be presented Wednesday morning.

0111

Glasgow IR Group @irglasgow.bsky.social · 08/04/2025

Yesterday, @ttsamurai.bsky.social has been busy at #ECIR2025, demoing FinPersona (w/ M Suzuki l, K Izumi, @javiersanzcruza.bsky.social, @richardmcc.bsky.social & @iadhounis.bsky.social). Great to see many people interested in FinTech LLM agents! Try Demo at: huggingface.co/spaces/TTsam...

062

Reposted by Glasgow IR Group

Jack McKechnie @jackmckechnie.bsky.social · 05/04/2025

Happy to share that our (w/ @grahammcdonald.bsky.social and @craigmacdonald.bsky.social) work "Measuring Hypothesis Testing Errors in the Evaluation of Retrieval Systems" has been accepted to #SIGIR2025 as a short paper! 🎉Exciting times! 🇮🇹 @ir_glasgow

0101

Glasgow IR Group @irglasgow.bsky.social · 03/04/2025

A large contingent of our staff & students @uofgcompsci.bsky.social 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!

0208

Reposted by Glasgow IR Group

Iadh Ounis @iadhounis.bsky.social · 18/03/2025

About a week left to apply to this postdoctoral position in Responsible AI/IR at @irglasgow.bsky.social Details are below. Deadline to apply: 26th March

067

Reposted by Glasgow IR Group

grahammcdonald.bsky.social @grahammcdonald.bsky.social · 17/03/2025

Watching @iadhounis.bsky.social give a Keynote talk about reproducibility and PyTerrier pipelines in Tsukuba x Glasgow workshop on the Frontiers of AI. @irglasgow.bsky.social

074

Reposted by Glasgow IR Group

grahammcdonald.bsky.social @grahammcdonald.bsky.social · 17/03/2025

We are recruiting a postdoc at @irglasgow.bsky.social to work on the Participatory Harm Auditing Workbenches and Methodologies (PHAWM) RAI UK project on Responsible AI. phawm.org Deadline: 26th March www.jobs.gla.ac.uk/job/research...

phawm.org

phawm

055

Reposted by Glasgow IR Group

Iadh Ounis @iadhounis.bsky.social · 26/02/2025

Come work with us in Glasgow! We are recruiting a postdoc at @uofgcompsci.bsky.social to work on understanding & measuring the harm of AI-driven IR approaches within various search contexts as part of the PHAWM project on Responsible AI. Deadline: 26/03 Details: www.jobs.gla.ac.uk/job/research...

jobs.gla.ac.uk

Research Associate

Job PurposeTo make a leading contribution to the Participatory Harm Auditing Workbenches and Methodologies (PHAWM) project. Working with Information Retrieval experts Prof Iadh Ounis, Prof Craig Ma...

066

Reposted by Glasgow IR Group

Zixuan Yi @zixuanyi.bsky.social · 23/01/2025

Happy to share that our paper entitled "A Multi-modal Large Language Model with Graph-of-Thought for Effective Recommendation" has been accepted by @naaclmeeting.bsky.social as a main conference paper, co-authored with @iadhounis.bsky.social. #NAACL2025

084

Reposted by Glasgow IR Group

Iadh Ounis @iadhounis.bsky.social · 23/01/2025

We have 2 tenure-track faculty openings in Machine Learning at @uofgcompsci.bsky.social; Applications are welcome in the general areas of AI & ML, which complement existing strengths in the @idaglasgow.bsky.social section. Details at: jobs.gla.ac.uk/job/lecturer-s… Deadline: 17 Feb 2025 at 23:45

jobs.gla.ac.uk

0105

Reposted by Glasgow IR Group

Zixuan Yi @zixuanyi.bsky.social · 14/01/2025

Happy to share that our paper entitled “Enhancing Recommender Systems: Deep Modality Alignment with Large Multi-Modal Encoders” has been accepted by ACM Transactions on Recommender Systems - joint work with Zijun Long, @iadhounis.bsky.social, @craigmacdonald.bsky.social and @richardmcc.bsky.social.

063

Reposted by Glasgow IR Group

Jack McKechnie @jackmckechnie.bsky.social · 07/01/2025

At 15:00 on 13th January, Francesco Luigi De Faveri from University of Padua will give an #IRTalk entitled "Preserving Privacy in Information Retrieval: Leveraging Differential Privacy Across Domains". Details: samoa.dcs.gla.ac.uk/events/viewt... @uofgcompsci.bsky.social @irglasgow.bsky.social

032

Reposted by Glasgow IR Group

IEEE ICDCS 2025 @icdcs2025.bsky.social · 08/01/2025

🚨A week to go (!) before the poster papers abstract deadline for #ICDCS2025🚨 Submissions should be in the form of papers in a short format that are best presented as posters. icdcs2025.icdcs.org/call-for-pos... Deadlines: 15 Jan (abstracts), 22 Jan (papers).

044

Reposted by Glasgow IR Group

Jack McKechnie @jackmckechnie.bsky.social · 06/01/2025

Today, Xiaoyu Zhang from Shandong University is giving an #IRTalk entitled "Enhancing Conversational Recommender Systems: Integrating Knowledge, Context, and Emotions". Details: samoa.dcs.gla.ac.uk/events/viewt... @uofgcompsci.bsky.social @irglasgow.bsky.social

011

Reposted by Glasgow IR Group

Jack McKechnie @jackmckechnie.bsky.social · 04/01/2025

At 15:00 on 6th January, Xiaoyu Zhang from Shandong University will give an #IRTalk entitled "Enhancing Conversational Recommender Systems: Integrating Knowledge, Context, and Emotions". Details: samoa.dcs.gla.ac.uk/events/viewt... @uofgcompsci.bsky.social @irglasgow.bsky.social

052

Reposted by Glasgow IR Group

craigmacdonald.bsky.social @craigmacdonald.bsky.social · 17/12/2024

Glad to share that our work (by Fanghzhen Tian and w/ @gdebasis.bsky.social and myself) "Is Relevance ‘Lost in Transmission’ from Retriever to Generator?" has been accepted to #ecir2025 as a full paper. See you next spring in Lucca 🙋🏻‍♂️ @irglasgow.bsky.social

0121