🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Search & recsys have historically drawn inspiration from language modeling, such as Word2vec for item embeddings, and GRUs and BERT for sequential models. LLMs are no different and have led to updates in model architecture, training data & paradigms, etc. eugeneyan.com/writing/recsys…

reliable keyword search is a basic expectation

I have a different understanding regarding the extent to which a dataset truly represents users' needs arxiv.org/abs/2212.02726 AI may be trained to answer questions that no one would ever ask, yet still achieve a high score.

Are We Solving a Well-Defined Problem? A Task-Centric Perspective on Recommendation Tasks Aixin Sun 孙爱欣 emphasizes the need for well-defined problem definitions that account for temporal dynamics, candidate item selection, and user interaction patterns. 📝 arxiv.org/abs/2503.21188

When peer review becomes pure review

A large contingent of our staff & students UofG Computing Science 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 UoG IDA Section, Glasgow CS

we spend billions training recommendation systems but cant even agree on what good recommendations actually means...feels like trying to optimize a compass without knowing which way is north 🧭 aimodels.fyi/papers/arxiv/a…

Paper got rejected despite the archive version received 6 citations during the review period. Probably this can be translated into 6 positive reviews from a different pool of reviewers :)

The interpretation seems to include some of its (or LLM's) own thoughts and extensions. As the author, I find these interpretations quite interesting, though I don't fully agree with them. aimodels.fyi/papers/arxiv/a…

This working paper extends my earlier invited paper published in the SIGWEB Newsletter.

In the year 2025, we published a paper praising BERT-based models. Yes, we know larger models exist. ieeexplore.ieee.org/abstract/docum…

🚀 Excited to announce #GENNEXT at #SIGIR2025 — the first workshop dedicated to next-gen IR & recommender systems with #agents, LLMs, and conversational AI! 📄 Papers will be indexed in CEUR 🗓️ Submission deadline: April 23, 2025 sigirgennext.github.io/GENNEXT-SIGIR-…

Seems a reasonable answer

30年没回西安了

I like Table 1 a lot arxiv.org/pdf/2410.22972 And thanks for citing my work :)

Asked ChatGPT to check grammar on a simple True/False Wooclap question. It said your statement was wrong and explained why. I literally just asked for a grammar check 😅

Text+image shall be read and understood together

Excited to share the Societal AI research agenda—a cross-disciplinary effort to align AI with human values and societal needs. Check out our white paper and podcast hosted by Gretchen Huizinga! Thanks to all contributors! #SocietalAI #ResponsibleAI #MicrosoftResearch