English

Lightweight Inference-Time Personalization for Frozen Knowledge Graph Embeddings

Artificial Intelligence 2025-12-30 v1

Abstract

Foundation models for knowledge graphs (KGs) achieve strong cohort-level performance in link prediction, yet fail to capture individual user preferences; a key disconnect between general relational reasoning and personalized ranking. We propose GatedBias, a lightweight inference-time personalization framework that adapts frozen KG embeddings to individual user contexts without retraining or compromising global accuracy. Our approach introduces structure-gated adaptation: profile-specific features combine with graph-derived binary gates to produce interpretable, per-entity biases, requiring only 300{\sim}300 trainable parameters. We evaluate GatedBias on two benchmark datasets (Amazon-Book and Last-FM), demonstrating statistically significant improvements in alignment metrics while preserving cohort performance. Counterfactual perturbation experiments validate causal responsiveness; entities benefiting from specific preference signals show 6--30×\times greater rank improvements when those signals are boosted. These results show that personalized adaptation of foundation models can be both parameter-efficient and causally verifiable, bridging general knowledge representations with individual user needs.

Keywords

Cite

@article{arxiv.2512.22398,
  title  = {Lightweight Inference-Time Personalization for Frozen Knowledge Graph Embeddings},
  author = {Ozan Oguztuzun and Cerag Oguztuzun},
  journal= {arXiv preprint arXiv:2512.22398},
  year   = {2025}
}
R2 v1 2026-07-01T08:42:14.716Z