High Precision Audience Expansion via Extreme Classification in a Two-Sided Marketplace

2026-02-16

AI Summary

This work addresses the critical challenge of efficiently retrieving high-intent booking regions in two-sided marketplaces like Airbnb, where such retrieval underpins subsequent high-cost ranking stages. The authors propose a geospatial retrieval method based on extreme classification, partitioning the globe into 25 million uniform rectangular grids and applying extreme classification—novelly adapted for geographic audience expansion—to enable large-scale multi-label region prediction. Compared to the prior deep Bayesian bandit approach, this method substantially improves both relevance and retrieval efficiency of candidate regions. Deployed in Airbnb’s production system, the solution effectively reduces unnecessary inventory lookups, thereby enhancing overall search performance and user experience.

Technology Category

• Machine Learning: Multi-class/Multi-label Learning & Extreme Classification
• Search and Optimization: Distributed Search
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Graph Algorithms and Modeling for the Web: Querying, indexing, and retrieval in Web-related graphs

Abstract

Airbnb search must balance a worldwide, highly varied supply of homes with guests whose location, amenity, style, and price expectations differ widely. Meeting those expectations hinges on an efficient retrieval stage that surfaces only the listings a guest might realistically book, before resource intensive ranking models are applied to determine the best results. Unlike many recommendation engines, our system faces a distinctive challenge, location retrieval, that sits upstream of ranking and determines which geographic areas are queried in order to filter inventory to a candidate set. The preexisting approach employs a deep bayesian bandit based system to predict a rectangular retrieval bounds area that can be used for filtering. The purpose of this paper is to demonstrate the methodology, challenges, and impact of rearchitecting search to retrieve from the subset of most bookable high precision rectangular map cells defined by dividing the world into 25M uniform cells.

Problem

• audience expansion
• extreme classification
• two-sided marketplace
• location retrieval
• high precision retrieval

Innovation

• Extreme Classification
• Audience Expansion
• Location Retrieval
• Two-Sided Marketplace
• High Precision Retrieval