English static mirror for SEO/GEO · AI-assisted translation · Read Chinese original

Learning To Rank Diversely at Airbnb (CIKM 2023)

Forum topic · 小凯 · 2026-07-05

Summary

This entry describes a CIKM 2023 paper from Airbnb titled "Learning To Rank Diversely," which addresses diversification in learning-to-rank (LTR) systems for large-scale search and recommendation. In Airbnb's marketplace setting, a ranking list that over-concentrates on a single type of listing can hurt long-tail exposure, user satisfaction, and booking outcomes, motivating ranking approaches that balance relevance with result diversity. The work sits at the intersection of information retrieval, recommendation, and industrial ranking-system design, where efficiency, scalability, and understanding user intent remain core challenges. The forum entry contextualizes the paper within the broader evolution of neural ranking — from BM25 and dense bi-encoders to cross-encoder re-rankers and LLM-era agentic retrieval — and highlights recurring engineering constraints such as latency budgets, inference cost, evaluation credibility, and safety. It also provides checklists for data governance, latency, quality, security, and cost when translating paper-level results into production systems. Readers should consult the original PDF via the ACM DL link for exact methods, experimental numbers, and conclusions, as this entry is a structured overview based on the abstract and metadata.

Learning To Rank Diversely at Airbnb (CIKM 2023)

Paper: Learning To Rank Diversely At Airbnb, CIKM 2023 Source: https://dl.acm.org/doi/abs/10.1145/3583780.3614692 Type: Academic paper (CIKM 2023, Verticals track)

Key points

  • The paper tackles diversified learning to rank (LTR) at Airbnb, targeting the tension between ranking purely by relevance and producing diverse result lists in a marketplace search setting.
  • It belongs to the broader ranking research area, where large-scale search/recommendation systems face challenges in efficiency, scalability, and user-intent understanding.
  • In industrial systems, overly homogeneous rankings can reduce long-tail item exposure and user satisfaction, motivating diversity-aware objectives alongside relevance.
  • Context within modern ranking research

    The forum entry situates this work in the evolution of neural information retrieval:

  • From BM25 to BERT-style cross-encoders, dense bi-encoder retrieval, late interaction, and now generative retrieval / LLM agents.
  • Classic search stacks act as a funnel: recall for coverage, fine ranking for discrimination, generation for presentation. LLM-era systems add new variables: inference budget and action space (whether to retrieve, how often, which tools to call).
  • On the recommendation side, the core tension is sparse user behavior, huge item catalogs, and multi-objective business goals; LLMs add semantic priors and cold-start help but raise online inference cost and hallucination risk.
  • Engineering deployment checklist (from the entry)

    | Area | Question | Suggestion | |------|----------|------------| | Data | PII in training/index? Versioning? | Partitioned indexes, anonymization, rollback-capable embedding versions | | Latency | p99 budget? Retrieval steps? | Cascade + early stopping, hot-query caching, async re-ranking | | Quality | Does offline gain convert to online CTR/satisfaction? | Interleaving experiments, human audit samples, citation checks | | Safety | Does open retrieval introduce poisoning/bias? | Source whitelists, adversarial detection, output filtering | | Cost | Per-query tokens and GPU usage? | Small-model routing, distillation, hybrid sparse+dense |

    Takeaways for search / rec / personalization

    1. Architecture: cascade of retrieval + re-ranking + generation remains mainstream, but agentic paradigms make retrieval strategy itself learnable. 2. Data: high-quality instruction data and click/session logs both matter; synthetic data needs leakage and distribution-shift controls. 3. Evaluation: the gap between offline metrics and online satisfaction is widening; LLM-as-judge needs cross-validation with human evaluation. 4. Product: latency, cost, explainability, and safety are hard production constraints — academic benchmarks alone are not enough.

    Related entries

  • An interpretable ensemble of graph and language models for improving search
  • Applying Deep Learning to Ads Conversion Prediction in Last Mile Delivery
  • Automated Query-Product Relevance Labeling using Large Language Models
  • Behavior Modeling Space Reconstruction for E-Commerce Search (Jan 2025)
  • Better to Ask in English: Cross-Lingual Evaluation of Large Language Models
  • Reference

  • Original paper: Learning To Rank Diversely At Airbnb, CIKM 2023 — ACM DL
> Note: Quantitative results should be verified against the original PDF; this entry is a structured overview based on the abstract and public metadata.

Tags

#learning-to-rank#airbnb#search-ranking#diversity#recommender-systems#information-retrieval#cikm-2023#industrial-applications

This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/178209021