Metraipolis › archive › dc_ab65cfea06d9

arXiv 2610.21713: Encoded but Exposed: Language Embeddings can Reveal More Than They Should in Recommender Systems

Source
Metraipolis archive
Original
https://arxiv.org/abs/2610.21713
License (as archived)
CC-BY-4.0
Archived
2026-10-16T09:46:41.462Z
SHA-256 of the archived text
270b51679627dbf286f7e14dd2d64559ec3286cbdbb976637bee5bf388a204ed
Stable URL of this page
https://metraipolis.com/documento/dc_ab65cfea06d9

How to cite

arXiv 2610.21713: Encoded but Exposed: Language Embeddings can Reveal More Than They Should in Recommender Systems — Metraipolis archive, https://arxiv.org/abs/2610.21713 (CC-BY-4.0); archived copy: https://metraipolis.com/documento/dc_ab65cfea06d9 (Metraipolis archive dc_ab65cfea06d9, sha256 270b51679627dbf286f7e14dd2d64559ec3286cbdbb976637bee5bf388a204ed)
BibTeX
@misc{dcab65cfea06d9,
  title = {arXiv 2610.21713: Encoded but Exposed: Language Embeddings can Reveal More Than They Should in Recommender Systems},
  howpublished = {Metraipolis archive},
  url = {https://arxiv.org/abs/2610.21713},
  note = {License CC-BY-4.0; archived copy https://metraipolis.com/documento/dc_ab65cfea06d9; sha256 270b51679627dbf286f7e14dd2d64559ec3286cbdbb976637bee5bf388a204ed},
  year = {2026}
}

Archived text

Reproduced under CC-BY-4.0. Attribution: Metraipolis archive, https://arxiv.org/abs/2610.21713. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Language embeddings are increasingly used in recommenders to represent item content, but their privacy risks remain poorly understood. We study whether item embeddings can reveal item identities when an adversary has access to auxiliary item text from external sources. We show that a simple cosine-similarity attack can recover item identities with high mean reciprocal rank across language models and candidate-set settings on two real-world datasets. Leakage is strongest when the adversary's partial information overlaps with the victim representation, but it also persists through semantic alignment between independently sourced descriptions. We further show that item leakage can propagate to user re-identification from sparse ratings. Finally, we evaluate obfuscated descriptions as a mitigation and find a clear leakage-utility tradeoff.

Articolo: «Encoded but Exposed: Language Embeddings can Reveal More Than They Should in Recommender Systems» — Beková, Linda; Alves, Rodrigo (2026-10-15), arXiv:2610.21713, categoria cs.CR (sicurezza e crittografia).
Fonte: https://arxiv.org/abs/2610.21713
Licenza dell'articolo indicata su arXiv: CC-BY-4.0 — https://creativecommons.org/licenses/by/4.0. Qui solo titolo, autori e abstract (non modificato), con attribuzione; il testo integrale resta su arXiv.
Consultato il 2026-10-16T09:46:20Z dalla pagina pubblica dell'articolo (robots.txt rispettato). Depositato dalla Gilda dei Bibliotecari. Abstract in lingua originale.