Papers
arxiv:2611.23236

PGPO: Prototype-Guided Progressive Obfuscation for Privacy-Preserving LLM-Enhanced Recommendation

Published on Aug 26
· Submitted by
Shengxiang Lin
on Nov 19
Authors:
,
,
,
,
,

Abstract

A prototype-guided progressive obfuscation method preserves relative item relations while perturbing high-influence semantic embeddings to defend against inversion attacks in LLM-enhanced recommender systems.

LLM-enhanced recommender systems encode item side information into semantic embeddings, but these embeddings may leak sensitive textual attributes through inversion attacks. Existing input-side defenses rely on trusted collection or local token perturbation, often disrupting the inter-item semantic topology needed for recommendation. To address it, we introduce relational fidelity as the key objective, requiring semantic shift for privacy while preserving relative item relations for utility. Based on this principle, we propose prototype-guided progressive obfuscation (PGPO), which learns topology-aware obfuscation anchors for high-influence prototype words and propagates them through a static semantic graph. PGPO is optimized with GRPO using individual dissimilarity, structural consistency, and semantic constraint rewards. Experiments on two benchmarks show strong inversion defense with near-upper-bound recommendation performance. The code is available at https://github.com/Shengxiang-Lin/PGPO.

Community

Paper author Paper submitter
•
edited about 6 hours ago

Keywords: LLM-Enhanced Recommendation; Privacy-Preserving; Reinforcement Learning

TL;DR: PGPO protects item side information in LLM-enhanced recommendation by inducing semantic shift while preserving relational topology, reducing embedding inversion risks with minimal utility loss.

Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing.

Sign up or log in to comment

Get this paper in your agent:

hf papers read 2611.23236
Don't have the latest CLI?
curl -LsSf https://hf.co/cli/install.sh | bash

Models citing this paper 1

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2611.23236 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2611.23236 in a Space README.md to link it from this page.

Collections including this paper 0

No Collection including this paper

Add this paper to a collection to link it from this page.