AtomRec: Evolving Atomic Memory for Agentic Recommendation

2026-09-04

AI Summary

本文提出AtomRec，通过构建和演化用户与项目间的结构化原子记忆单元及语义链接，解决现有推荐系统中偏好细节丢失和证据难解释的问题。

Abstract

Agentic recommender systems use large language models to maintain semantic memory and support evidence-aware recommendation. However, existing memory mechanisms often compress user and item information into coarse summaries and connect them with scalar collaborative links, making it difficult to preserve fine-grained preference stages or retrieve interpretable evidence as user interests evolve. We propose \textsc{AtomRec}, an agentic recommender with evolving atomic collaborative memory. \textsc{AtomRec} represents user and item memories as structured atomic units, builds semantic links across related memories, and evolves related historical fields when new interactions arrive. During recommendation, it retrieves linked memories as multi-hop evidence paths rather than isolated neighbor summaries, allowing collaborative signals to support grounded ranking. Experiments on four public benchmarks show that \textsc{AtomRec} consistently outperforms state-of-the-art agentic and memory-augmented baselines, with around 8.5\% average relative improvement across metrics.

Problem

• agentic recommender systems
• semantic memory
• fine-grained preference
• interpretable evidence

Innovation

• Evolving Atomic Memory
• Semantic Links
• Multi-hop Evidence Paths
• Agentic Recommendation