Search ranking for a stays marketplace
Inspired by publicly reported work at Airbnb
Rank hotel and stay results for each search so guests find and book faster, using learning to rank.
The brief
Airbnb has described how its search ranking grew from a random order to gradient-boosted models trained on searches that led to bookings. Build that journey on open hotel-search data: start with a simple ranking, train learning-to-rank models on clicks and bookings, and run the ranking behind a search page. Good looks like: NDCG against simple baselines, a feature study (price, reviews, distance, personal history), a search demo, and a plan for an online A/B test.
Read the case first
What Airbnb has said in public about this kind of work, on its own pages:
- Machine Learning-Powered Search Ranking of Airbnb Experiences (Airbnb Engineering) ↗
- Learning Market Dynamics for Optimal Pricing (Airbnb Engineering) ↗
Make it your own
Rank something other than stays: tours and activities, coaching centres or event venues. Decide what a good result means for that market (distance, budget, language) and add one fairness rule for new listings.
AITeky has no link with Airbnb, which has not reviewed or endorsed this brief. The brief only draws on what the company has published, uses open data rather than its data, and is a learning exercise: don't present your project as Airbnb's work or system.
What you will build
- Explore searches, impressions, clicks and bookings
- Build baselines (price, rating, popularity)
- Train LambdaMART or LightGBM rankers
- Evaluate with NDCG@5 by search type
- Add personalisation from a guest's past behaviour
- Build a search page demo and write an A/B test plan
Free, open data to build it with
Sign up free to get the data and links
It takes a minute with your email. AiTeky Pass (₹999/month or ₹9,999/year) is free for your first 30 days.