trends.barnas.net

task · first seen 6 Oct 2026

recommender systems

Also: recommendation system, job recommendation system, recommendation algorithms

Methods that predict user preferences for items based on past interactions and item features.

Recommender systems predict which items a user will want from past interactions and item features. They have been a core area of applied machine learning for decades. The sources are a broad mix rather than one new idea. Several papers use LLM agents to automate recommender development through recursive self-improvement, including RecRSI, DualEvolve and ranking-scored trials. Others cover personal-agent mediated recommendation across platforms, Meta's CLEAR system with editable natural-language interests, the KuaiRecV2 benchmark, harmful-content propagation, and travel recommenders. Hobby projects such as a job-recommendation repo and a Show HN web-swiping feed also appear. The spread reflects ongoing activity in a mature field, not a newly emerging concept.

The umbrella term itself is not new. The sub-threads inside it may be worth tracking separately: LLM agents that self-improve recommenders, and user-controllable, legible preference layers.

Analyst note (established), 17 Nov 2026.

Mentions per week

0319 May11 Aug10 NovPapersCodeDiscussion

PapersCodeDiscussion

Rank 47 on 14 Nov 2026 · 11 mentions in 28 days · 11 independent groups · papers → code → discussion · first seen 6 weeks ago · acceleration +0.81

Earliest sources

Latest sources

Mentioned together with

personal agentmultiple instance learningmembership and identity inference attackmulti-agent LLM systemrecursive self-improvementLLM-driven scientific discoveryLLM agent self-evolutioncontinual learningnatural-language user profiles