Metraipolis › archive › dc_7aa0f15644b2

Keeping the Index Open: The Recommendation-Side Cost of Shared Search and Recommendation — arXiv:2608.24079

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.24079
License (as archived)
CC0-1.0
Archived
2026-10-18T06:36:42.121Z
SHA-256 of the archived text
92a403f2055dd490ccffce87be518a1d5c2c1c9ab266b9a61ef947c1a92d2975
Stable URL of this page
https://metraipolis.com/documento/dc_7aa0f15644b2

How to cite

Keeping the Index Open: The Recommendation-Side Cost of Shared Search and Recommendation — arXiv:2608.24079 — arXiv (en), https://arxiv.org/abs/2608.24079 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7aa0f15644b2 (Metraipolis archive dc_7aa0f15644b2, sha256 92a403f2055dd490ccffce87be518a1d5c2c1c9ab266b9a61ef947c1a92d2975)
BibTeX
@misc{dc7aa0f15644b2,
  title = {Keeping the Index Open: The Recommendation-Side Cost of Shared Search and Recommendation — arXiv:2608.24079},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.24079},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7aa0f15644b2; sha256 92a403f2055dd490ccffce87be518a1d5c2c1c9ab266b9a61ef947c1a92d2975},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.24079. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

A shared search-and-recommendation index must score new items from features alone because search has no exploration slot. In a public log covering both surfaces over one catalog, $38.6\%$ of held-out query-search impressions show an item never previously shown or visited. For user-cold engagements, the feature-based tower serves this demand without measurable loss against $99$ sampled negatives ($0.9595$ Recall@20 versus $0.9510$ warm). A lexical baseline reaches similar parity, while a full-catalog check remains statistically undecided. Dual-encoder retrieval therefore keeps the index \emph{open} to new items, unlike an ID-softmax recommender that requires retraining. We price this openness on recommendation against six sequential baselines, each retrained and tuned through five rounds on corrected targets. A float32 timestamp bug had reordered leave-one-out targets for $19.7\%$ of users. On MovieLens-1M, warm accuracy trails the strongest retrained baseline by $5.2\%$ Recall@20 and $11.4\%$ NDCG@20. On MIND, the gap narrows to $0.8$--$3.6\%$ relative to the five strongest baselines, though the model ranks sixth of seven. Under strict zero-leakage cold-start evaluation, the content tower achieves $0.172 \pm 0.006$ Recall@20, $1.4\times$ the strongest retrained dedicated method ($0.124 \pm 0.007$) and $3\times$ a training-free floor, without cold-specific training. Exact full-softmax training raises Recall@20 by $54\%$ on MIND-small and $6.9\%$ on MovieLens-1M over sampled InfoNCE, but recomputes the full catalog each step and exhausts accelerator memory at $240$K items. Approximate nearest-neighbor search explains none of the remaining gap, serving cost does not regress against ID-softmax retrieval, and a history-window sweep explains half the post-recipe remainder. Exact-quality training at catalog scale remains the open problem.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-26 in cs.LG).
Categorie arXiv: Information Retrieval (cs.IR) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2608.24079 — 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 2026-10-18T06:28:52.336Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-08-26?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).