AdaRec: LLMs that explain and adapt your recommendations

Kari Jaaskelainen

11 Nov 2025 — 1 min read

Meet AdaRec, a new way to build personalized recommendations with large language models that works even when you have very little data.

Instead of hand-crafted features, AdaRec uses narrative profiling—it turns a person’s interactions into plain-language mini-stories the model can understand and explain.

Peer patterns (horizontal alignment): learns from similar users’ behavior.
Why it fits (vertical attribution): highlights the key reasons behind a user’s preferences.

Across real e‑commerce datasets, AdaRec beat traditional ML and LLM baselines by up to 8% in few-shot tests. In zero-shot settings, it surpassed expert-designed profiles by up to 19%, making it especially strong for long‑tail items with minimal interaction data. Lightweight fine‑tuning on AdaRec’s own synthetic data matched fully fine‑tuned models—saving time and compute.

Fewer labels, clearer reasoning, faster adaptation.

Paper by Meiyun Wang and Charin Polpanumas. Read more: http://arxiv.org/abs/2511.07166v1

Paper: http://arxiv.org/abs/2511.07166v1

Register: https://www.AiFeta.com

#RecommenderSystems #LLM #AI #Personalization #Ecommerce #NLP #FewShot #ZeroShot

Automating GDPR Compliance: A Roadmap for Companies and Law Firms

GDPR compliance is more than checkboxes. A new roadmap from the Privatech project shows how automation and machine learning can help companies and law firms assess—and even generate—privacy compliance. * Shift the focus to data processors’ real workflows: drafting policies, mapping data uses, documenting decisions. * Break compliance into machine-ready

FPGAs for Faster, Leaner Deep Learning: A Review of CNN Accelerators

Deep learning drives image search, robots, and medical scans. Most systems lean on CPUs and GPUs. This review asks: what if we run convolutional neural networks (CNNs) on FPGAs—reconfigurable chips you can tailor to the model? * Why FPGAs: custom dataflows, low latency, and strong energy efficiency—great for cameras,

Dynamic-K: Recommendations That Know When to Stop

Most apps show a fixed number of “top” items—say 10 movies or 20 products—assuming there are always enough good options. But that’s not always true: sometimes there are few relevant items, or some users are extra picky. The result? Filler recommendations. Dynamic-K flips the script. Instead of

Teaching chatbots to stop contradicting themselves (DECODE)

Teaching chatbots to stop contradicting themselves Ever had a bot say one thing, then the opposite a few turns later? This study introduces DECODE—a new task and dataset for spotting contradictions in everyday conversations, drawn from both human-human and human-bot chats. * New data beats existing natural language inference (NLI)