Where menu data comes from

Restaurants and locations

Restaurant names, addresses, categories and map positions come from open place datasets, mainly the Overture Maps Foundation, plus places added by TasteRex users and by restaurants themselves. Open data is used under its licences; full attribution is in the app under Settings → Data sources.

A chain is listed as one restaurant with many locations, so a dish like a chain’s signature burger has one shared record and every location’s ratings count toward it.

Dishes

Dish names come from three places:

1. The restaurant’s own published menu. Our crawler reads menu pages on the restaurant’s public website, or a menu the restaurant has posted publicly (such as a PDF or a photo of the menu). AI then reads the menu and sorts each dish into a cuisine, a category (like Burgers or Ramen) and a specialty.
2. Diners. If a dish is missing, a diner can add it while rating it.
3. Restaurants. Once a restaurant claims its page, it can add, rename and retire dishes itself.

Uncertain classifications go to a person for review before they’re published.

What we publish, and what we don’t

Dietary information

TasteRex never guesses whether a dish is gluten-free, vegan, nut-free or anything else, even if a menu description suggests it. A description can’t tell you about a shared fryer or a dusted bun; only the kitchen knows. Dietary details appear on a dish only when the restaurant itself has supplied them, and we label them as coming from the restaurant. Even then, always confirm allergens with the restaurant before ordering.

“Unverified listing”

A dish marked unverified listing came from a menu we read, and no one has rated it yet. It may be misspelled, duplicated or no longer served. The label goes away once diners rate the dish, or when the restaurant confirms its menu after claiming its page.

Separately, a rating can be a verified dine, meaning the diner’s phone confirmed they were at the restaurant. Ratings without that check are still welcome; they just don’t carry the badge.

Our crawler

Our crawler identifies itself as TasteRexBot (see tasterex.com/bot). It reads public pages only, never signs in or gets around paywalls, follows each site’s robots.txt, makes one request at a time per site, and pauses between requests. Where a site can’t be read, we record the reason and leave the menu to be filled by the restaurant or by diners.

Keeping it accurate

• Diners can suggest a correction, report a dish that’s not on the menu, or report a duplicate from the dish’s menu in the app.
• Restaurants can claim their page for free and edit their details, menu, photos and dietary information directly. Edits made by a restaurant are protected: our next automatic update will never overwrite them.
• Opting out. A restaurant that doesn’t want its menu read automatically can block TasteRexBot in robots.txt, or email restaurants@tasterex.com. The restaurant stays listed so diners can still rate it, but we’ll stop reading its website.