Metraipolis › archive › dc_2ff34269c95b

FairDiff: Mitigating the Self-Reinforcing Matthew Effect in Diffusion Recommender Models — arXiv:2609.36671

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.36671
License (as archived)
CC0-1.0
Archived
2026-10-07T16:22:52.548Z
SHA-256 of the archived text
0a4170c45386f77ff052633d520f0fbfed6aace446e758cf92dd2fda4dea0ed0
Stable URL of this page
https://metraipolis.com/documento/dc_2ff34269c95b

How to cite

FairDiff: Mitigating the Self-Reinforcing Matthew Effect in Diffusion Recommender Models — arXiv:2609.36671 — arXiv (en), https://arxiv.org/abs/2609.36671 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_2ff34269c95b (Metraipolis archive dc_2ff34269c95b, sha256 0a4170c45386f77ff052633d520f0fbfed6aace446e758cf92dd2fda4dea0ed0)
BibTeX
@misc{dc2ff34269c95b,
  title = {FairDiff: Mitigating the Self-Reinforcing Matthew Effect in Diffusion Recommender Models — arXiv:2609.36671},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.36671},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_2ff34269c95b; sha256 0a4170c45386f77ff052633d520f0fbfed6aace446e758cf92dd2fda4dea0ed0},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.36671. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

While the "Matthew Effect" and filter bubbles are widely recognized outcome-level biases in recommender systems, we reveal that Diffusion Recommender Models (DRMs) uniquely compound this issue through their generative dynamics. Rather than merely inheriting data imbalances, DRMs trigger a self-reinforcing amplification of popularity bias. We identify that this phenomenon is driven by two compounding mechanisms. First, while optimization loss is universally dominated by high-frequency items across recommenders, DRMs suffer from a unique structural prior mismatch during generation. Because the forward terminal distribution of long-tailed data deviates significantly from the standard Gaussian prior, reverse sampling trajectories inherently collapse toward high-density popular items, fundamentally suppressing niche item generation. To dismantle this self-reinforcing loop, we propose FairDiff, a plug-and-play fairness-aware diffusion framework. To overcome the popularity-dominated loss, we introduce Popularity Condition Guidance (PCG). Rather than altering the training objective, PCG acts as an inference-time distributional reweighting mechanism, mathematically reshaping the score-based gradient field to penalize high-popularity regions and guide trajectories toward niche semantics. Furthermore, we design a Semantic Calibration (SC) Module to bridge the prior mismatch, aligning the forward and reverse distributions via one-step optimal transport. Comprehensive evaluations demonstrate that FairDiff achieves state-of-the-art performance while effectively mitigating the self-reinforcing Matthew Effect, highlighting its value as a general framework for DRMs.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-30 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.36671 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-07T15:54:53.525Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-30?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).