The preprint will be available very soon! 📄 Looking forward to presenting our work and connecting with everyone at #CIKM2025 in Seoul this November. See you there! 🇰🇷 #RecommenderSystems #DeepLearning #MachineLearning

A huge thanks also to @LucioNarducci and @TommasoDiNoia for their trust in our vision and for their crucial support during the final refinement phase. This work was a true team effort. 🙏

I want to extend a special and heartfelt thank you to my main co-authors, @MattAttimonelli and @danilo_danese. This paper is a direct result of their brilliant insights and deep commitment. I am incredibly proud of what we have accomplished together! 🙌

The results are super promising! Our approach not only boosts recommendation accuracy but the embeddings can be decoded back into structured text. This confirms they have a deep, practical understanding of the multimodal content. ✅

Our solution: We use Large Vision-Language Models (LVLMs) to generate "multimodal-by-design" embeddings. 💡 This creates powerful, semantically rich representations without needing the complex fusion steps typical in MMRS. #MultimodalAI #LVLM

The core question: When a Multimodal Recommender System (MMRS) performs well, is it truly understanding the item's image and text, or are we just seeing the effect of increased model complexity? We decided to find out. 🧐 #RecSys

Thrilled to announce our paper, "Do Recommender Systems Really Leverage Multimodal Content?", has been accepted at #CIKM2025! 🔥 We investigate if the gains in Multimodal RecSys come from true understanding or just bigger models. A deep dive into what really works. (Thread 👇)

Wow, honored to be named an Excellent PC Member at @SIGIRConf! Glad my obsessive... I mean detailed... reviewing style paid off. 🤓✨ Huge thanks to the chairs for the recognition and congrats to all the other reviewers! #SIGIR #SIGIR2025 #Award #BestReviewer

📢Great news for #RecSys researchers! 📢 The @DaQuaMRec workshop CfP deadline is now **July 17**. One more week to submit your paper! Don't miss out.🚀 CC @dmalitesta @JannachDietmar @ale_melchiorre @Shahnawazgrewal @yubink_cs @JulianMcauley https://t.co/tfufUxaEao

Bad news from #RecSys2025 @ACMRecSys? 🥲 You still have time to submit your work to #DaQuaMRec 🚀 (**July 10, 2025**) Many topics in #multimodal #recommendation, such as #bias, #fairness, #missing #noisy data, modalities #misalignment, evaluation of data #quality, and more 🥰

🚀 CFP Alert! The Call for Papers for the 1st Workshop on Data Quality-Aware Multimodal Recommendation at #RecSys2025 is now online! 🎯 Focus: Data quality & bias in multimodal recommender systems 📅 Deadline: July 10, 2025 🌐 https://t.co/DFFoYULaJT @ACMRecSys #MultimodalAI

There is still plenty of opportunity to join the #RecSys2025 Challenge! Check out the details on their website: https://t.co/tnRpAoyWAb

Slides and code here: https://t.co/2ux54jmPd6

Huge thanks to all the attendees for the engaging and fruitful discussion! 🙏 Your questions and insights were fantastic and really pushed the conversation forward. Let's keep striving for more reproducible and replicable research in RS & IR! 🚀

And as always, our silent guardian Fagiolino kept a close eye on things! 👀 He really doesn't tolerate shaky experiments. 😉 Let's just say he ensures our methodology is as robust as possible! #ResearchHumor #FagiolinoWatch #NoShakyExperiments

Reproducibility isn't just a buzzword! It's the bedrock of rigorous, trustworthy, transparent, and accountable research & science. 🔬 Whether you're building recommendation engines or information retrieval systems, solid foundations are key! #TrustworthyAI #AccountableResearch

Building on our lecture at the ACM Europe Summer School, we expanded our scope to explore reproducibility and replicability across bothRecommender Systems and Information Retrieval. Turns out, these principles are vital regardless of the specific field! 🤝 #ResearchIntegrity

Super thrilled to share that @ntonellotto, @sciueferrara and I just wrapped up our tutorial at @ecir2025! 🎉 Topic? Reproducibility & Replicability in RecSys+IR. It was fantastic to be part of the conference! #reproducibility #replicability #ecir2025

The data preprocessing phase in recommender systems can be long and tedious. With 🧩DataRec, that’s no longer the case! “DataRec: A Python Library for Standardized and Reproducible Data Management in Recommender Systems” has been accepted for publication at SIGIR 2025. 🚀🚀

questa al mio paese si chiama TRUFFA @AGCOMunica (3/3)