Methodology

How the hotel data is collected, matched and checked each month, and which open data it builds on.

Sources

• Places to stay and their facts: publicly available web listings and booking sites.
• Prices: nightly offers from a fixed list of known booking sites and from the hotels' own websites, captured monthly. Rates are indicative nightly offers captured for the search dates at the time of capture, not contracted rates.
• Contact details: as published by the properties, including publicly listed business email addresses found on the hotel's own website and official pages.
• Building footprints: OpenStreetMap and Overture Maps, matched to each hotel's point.

The monthly cycle

1. Collect the current listings and offers.
2. Match each listing to the existing place, keeping its stable ld_id.
3. Match each hotel to a building footprint and grade the match.
4. Compare the release with the previous one to produce the change history (history and changes).

Quality rules

• Match tiers: list matching scores name (70%) and city (30%) similarity, with a small bonus for places typed as hotels, and assigns high, medium, low or no_match (list matching).
• Polygon grades: each footprint match is graded A (highest confidence) to E (lowest), and its source is recorded as overture_contains, osm_name or overture_nearest. Grades A–B suit automated use; C–E deserve a review (building polygons).
• Gaps are gaps: a place missing from one release is treated as a data gap, not a closure. A closure is recorded only when the place is marked closed.
• Published facts only: contact fields are left empty when the property does not publish them.

Attribution

Building footprints are derived from open data and keep their original licences and attribution requirements:

• © OpenStreetMap contributors, available under the Open Database License (ODbL). openstreetmap.org/copyright
• Contains data from the Overture Maps Foundation, used under the licences listed in its attribution guidance.

The data licence explains how these open licences apply to delivered data.