yelp_com
Yelp business directory data, including businesses, locations, hierarchical categories, and associated image metadata. business-reviews local-search location-data
README
This database contains Yelp-derived business listings, geographic classifications, category hierarchies, and locally stored business images. The sample data covers locations including Buenos Aires, Argentina, and Toronto, Canada. Source values are retained in a form that may require application-level normalization.
Business listings
The business table stores the primary listing records, including business names and Yelp-style slugs. The sample business record also contains an address, neighborhood, phone number, star-rating display text, coordinates, and a thumbnail image URL. Addresses may contain HTML markup, and ratings are preserved as display strings such as 4.5 star rating rather than normalized numeric values.
Business slugs are generally URL-friendly identifiers, for example le-moulin-de-la-fleur-buenos-aires. Latitude and longitude values are stored as text in the sample data, so applications should convert them when performing numeric or geographic operations. Image URLs in business records point to Yelp’s media domain and may represent a thumbnail or another requested image size.
Geographic organization
Geographic data is represented through regions, cities, and districts. Cities reference regions through an identifier, while districts are intended to provide more granular location information and reference a city through city_id.
The source data is not uniformly normalized: a district slug may contain a complete search URL rather than a simple slug, and the sample district has city_id set to 0. Consumers should therefore treat geographic links, slugs, and relationship identifiers as source data that may require validation.
Categories
The category tables represent a two-level hierarchy, with broad groupings such as Food and more specific categories such as Bagels. Category records use slugs for source-oriented identifiers, while second-level categories reference their parent grouping through an identifier. As with the district sample, a category relationship identifier may be 0, so applications should not assume that every reference resolves to an existing parent record.
Business images
business_image stores image assets associated with businesses through business_id. It preserves the original source URL and content type, along with a content hash, file size, local storage path, and optional title.
The content hash and derived path support local asset storage and duplicate detection. Image URLs in the source data point to Yelp’s media domain and may refer to different image sizes or resolutions. The local path is organized from the content hash and can be used to locate the stored asset.
Data considerations
Records include ingestion-style identifiers and timestamps in the supplied data, allowing consumers to distinguish records and track when they were captured. Source values should generally be treated as strings unless an application explicitly normalizes dates, ratings, coordinates, URLs, or HTML-formatted addresses.
Relationship fields such as region_id, city_id, category_1_id, category_2_id, and business_id are intended to connect records, but the sample data includes zero-valued or otherwise unresolved references. Applications should validate these relationships before relying on them for joins or navigation.
Database Stats:
Database APIs:
- /business_1/ Business 1
-
/uk_restaurants_1/ UK Restaurants
Main: /main/list/