Metraipolis › archive › dc_3cd5b5ef0652
Multimedia Asset Personalization via Multimodal Embeddings at Netflix — arXiv:2608.18322
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.18322
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-18T14:03:40.123Z
- SHA-256 of the archived text
3549c978299b32c05f7b47c66de14461f0becb17d77ff7ecca1728ba2c25a0ed- Stable URL of this page
https://metraipolis.com/documento/dc_3cd5b5ef0652
How to cite
Multimedia Asset Personalization via Multimodal Embeddings at Netflix — arXiv:2608.18322 — arXiv (en), https://arxiv.org/abs/2608.18322 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_3cd5b5ef0652 (Metraipolis archive dc_3cd5b5ef0652, sha256 3549c978299b32c05f7b47c66de14461f0becb17d77ff7ecca1728ba2c25a0ed)
BibTeX
@misc{dc3cd5b5ef0652,
title = {Multimedia Asset Personalization via Multimodal Embeddings at Netflix — arXiv:2608.18322},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.18322},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_3cd5b5ef0652; sha256 3549c978299b32c05f7b47c66de14461f0becb17d77ff7ecca1728ba2c25a0ed},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.18322. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Personalized promotional assets, namely artwork images and video preview clips, are critical to content discovery on Netflix. Traditional models for asset selection rely on ID-based interaction history, leaving them blind to asset content and unable to serve newly launched titles and assets. We describe how multimodal embeddings reshaped production systems at Netflix and report transferable lessons for practitioners adopting foundation-model embeddings into recommender systems. First, pretrained image embeddings unlock cross-title, cross-canvas knowledge transfer. Augmenting a two-tower model with CLIP image embeddings lets a single model serve all five Netflix artwork canvas types, replacing five separately trained per-canvas models and substantially improving cold-start performance. A lightweight extension reuses CLIP's joint text-image space to make artwork personalization query-aware in search. Second, multimodality decisively beats any single modality for video preview personalization. We describe MediaFM, our in-house tri-modal foundation model trained on a large-scale corpus of shots from the Netflix show catalog, fusing visual (SeqCLIP), audio (wav2vec 2.0), and timed-text signals; adopted for video preview personalization, it outperforms strong visual-only baselines both offline and in online A/B tests. Third, a simple offline proxy task whose performance correlates with online outcomes can accelerate the experimentation and productization cycle. Predicting the popularity-based winner from embeddings alone ranks embedding models and versions, pruning the choice space before any end-to-end integration or A/B test; it now gates every new MediaFM checkpoint. We also share the production engineering decisions (shared embedding infrastructure, low-latency serving, cheap screening) that made these deployments viable, along with the design tradeoffs and failure modes we encountered. Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-08-20 in cs.IR). Categorie arXiv: Information Retrieval (cs.IR). Pagina del lavoro: https://arxiv.org/abs/2608.18322 — 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 2026-10-18T13:59:46.165Z dalla pagina https://arxiv.org/catchup/cs.IR/2026-08-20?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).