title: Amazon parent: Data Model
The Amazon reviews data set consists of user-provided reviews and ratings for a variety of products.
Currently we import the ratings-only data from the Books segment of the 2014 data set.
If you use this data, cite the paper(s) documented on the data set web site.
Imported data lives in the
az schema. The source files are not automatically downloaded.
The import is controlled by the following DVC steps:
az-schema.sql to set up the base schema.
: Import raw BookCrossing ratings from
az-index.sql to index the rating data and integrate with book data.
The raw rating data, with invalid characters cleaned up, is in the
az.raw_ratings table, with
the following columns:
user_key : The alphanumeric user identifier.
asin : The Amazon identification number for the product; for a book with an ISBN, this is the ISBN.
rating : The book rating. The ratings are on a 1-5 scale.
rating_time : The rating timestamp.
We extract the following tables for Amazon ratings:
: Mapping from Amazon's alphanumeric user identifiers to numeric user IDs.
: Rating values suitable for LensKit use, with numeric user and item identifiers. The ratings are
pre-clustered, so the book IDs refer to book clusters and not individual ISBNs or editions. This table has the following columns: `user_id` : The user ID. `book_id` : The [book code](ids.html#book-codes) for this book; the cluster identifier if available, or the ISBN-based book code if this book is not in a cluster. `rating` : The rating value; if the user has rated multiple books in a cluster, the median value is reported. `nactions` : The number of book actions this user performed on this book. Equivalent to the number of books in the cluster that the user has rated.
Press p or to see the previous file or, n or to see the next file