Where the data comes from.

Sources

Pipeline

1. Fetch. A bounded Overpass query retrieves named Singapore food-and-drink amenities. data.gov.sg download endpoints provide the current official GeoJSON files.
2. Normalise. Names, categories, coordinates, cuisine tags, websites, phones and accessibility fields are converted to one conservative schema. Unknown data stays unknown.
3. Enrich. URA polygons assign official planning areas. The nearest LTA station exit is calculated by geographic distance. HPB and NEA records are merged only when both name and location meet strict thresholds.
4. Validate. The refresh is rejected if coverage drops below expected floors, coordinates leave Singapore, identifiers collide, source attribution disappears, or official enrichments fall unexpectedly.
5. Publish atomically. A complete validated file replaces the live dataset in one operation, so a partial scrape cannot corrupt the site.

Ranking

Search results rank exact and partial name matches first, then cuisine and location matches, then listings with more useful details. There are no paid placements or user-rating scores.

Important limitations

• OpenStreetMap is community maintained and may be incomplete or out of date.
• Opening-hour strings are shown from source data and may not cover public-holiday exceptions.
• “Halal-tagged,” vegetarian, vegan, wheelchair, delivery and takeaway fields are source tags—not certifications or guarantees.
• HPB participation and NEA stall counts reflect their respective published datasets, not a real-time venue status check.
• Nearest MRT distance is measured to a recorded station exit in a straight line, not walking distance.

Attribution and reuse

OpenStreetMap data is © OpenStreetMap contributors and available under the Open Database Licence. Singapore government datasets are used under the Singapore Open Data Licence. DineCity preserves source links at listing level and in this methodology page.