AEAirbnb EngineeringAI & MLPro

CASTLE: training natural language search before the first real query

How we addressed the cold start problem for Airbnb’s natural language search by jointly generating synthetic training data — queries and relevance labels — from de identified booking and search signals. By : Wendy Ran Wei , Hao Li , Weiwei Guo , Xiaowei Liu , Xueyin Chen , Chun How Tan , Malay Haldar , Soumyadip…

Wendy Ran Wei 7 min readOriginal

How we addressed the cold-start problem for Airbnb’s natural language search by jointly generating synthetic training data — queries and relevance labels — from de-identified booking and search signals.

View across the lake in St. James’s Park, London, framed by golden autumn willow trees, with the ornate towers and domes of Horse Guards building visible in the background and swans and waterfowl swimming on the water with a fountain spraying in the center.

By: Wendy Ran Wei, Hao Li, Weiwei Guo, Xiaowei Liu, Xueyin Chen, Chun How Tan, Malay Haldar, Soumyadip Banerjee, Kedar Bellare, Huiji Gao, Sanjeev Katariya, and Stephanie Moyerman

Introduction

Keep reading Airbnb Engineering with Pro

This source is part of the Pro library. Unlock every premium source, AI translation and custom feeds.

Unlock with ProI already have Pro
#LLM#AI#Machine Learning#Data#technology
AE

Originally published on

Airbnb Engineering

Read original

Keep reading