Personal agent improves recommendations using cross platform user history
Personal-Agent Mediated Recommendation with Cross-Platform User History
Artificial IntelligenceMachine Learning
Summary
Recommendation systems usually suggest items based only on activity within one website or app. This paper explores using a personal AI assistant that can combine your activity across different platforms to make better recommendations. The challenge is balancing this cross-platform input with the original system's suggestions to avoid harmful changes. The authors propose a new method, PAMO, to train these agents to improve recommendations while minimizing mistakes. They test their approach on simulated and real data, showing it improves results compared to other methods.
What this means in practice
- •For mobile app developers: Build recommendation features that improve suggestions by combining user activity across multiple apps while balancing platform-specific rankings.
- •For online retail teams: Enhance personalized product suggestions by using customer behavior aggregated from multiple shopping sites mediated through a personal AI agent.
Authors
Yu Xia, Jiangfan Zhang, Jun Xiao, Julian McAuley, Xiangjun Fan
Abstract
Modern recommendation is shifting from platform-centric personalization toward user-governed personalization, where a personal LLM agent can act on the user's behalf across services. We formalize this emerging paradigm as Personal-Agent Mediated Recommendation: a platform recommender ranks a candidate set using platform-local information, and a personal agent uses user-authorized cross-platform history to mediate the resulting ranking and produce the final top-K slate. Such mediation is nontrivial: the platform ranking can encode strong population evidence that the personal agent cannot observe, so effective mediation must therefore balance beneficial rescues against harmful overrides. To study this trade-off, we introduce MediateRec, a benchmark that includes scalable proxy cross-platform environments and a real cross-platform test under a controlled platform-agent information boundary. To train the agent to use cross-platform history effectively, we further propose Personal Attribution Mediation Optimization (PAMO), which counterfactually masks that history to estimate personal mediation support and reallocates rank-aware advantage mass under a platform-relative value floor. We theoretically prove that PAMO preserves cutoff-level advantage mass and is locally optimal among first-order reallocations that preserve this mass without lowering average platform-relative value. Experiments on MediateRec show that personal-agent mediation enables meaningful platform corrections, yet even strong proprietary LLMs introduce non-negligible harmful overrides. PAMO consistently improves over matched outcome-only RL across seen and unseen target platforms and on the real cross-platform test, while achieving a better rescue-harm balance.