Amazon Review Data (2018)

Jianmo Ni, UCSD

Description

This Dataset is an updated version of the Amazon review dataset released in 2014. As in the previous version, this dataset includes reviews (ratings, text, helpfulness votes), product metadata (descriptions, category information, price, brand, and image features), and links (also viewed/also bought graphs). In addition, this version provides the following features:

• More reviews:
  • The total number of reviews is 233.1 million (142.8 million in 2014).
• Newer reviews:
  • Current data includes reviews in the range May 1996 - Oct 2018.
• Metadata:
  • We have added transaction metadata for each review shown on the review page. Such information includes:
    • Product information, e.g. color (white or black), size (large or small), package type (hardcover or electronics), etc.
    • Product images that are taken after the user received the product.
  • Added more detailed metadata of the product landing page. Such detailed information includes:
    • Bullet-point descriptions under product title.
    • Technical details table (attribute-value pairs).
    • Similar products table.
• More categories:
  • Includes 5 new product categories.

You can also download the review data from our previous datasets.

Amazon review (2014)

Amazon review (2013)

Citation

Please cite the following paper if you use the data in any way:

Justifying recommendations using distantly-labeled reviews and fined-grained aspects
Jianmo Ni, Jiacheng Li, Julian McAuley
Empirical Methods in Natural Language Processing (EMNLP), 2019
pdf

News

05/2021 We updated high resolution image urls to the metadata!

08/2020 We have updated the metadata and now it includes much less HTML/CSS code. Feel free to download the updated data!

Note

We provide a colab notebook that helps you parse and clean the data. For example:

• Load the metadata (e.g. as JSON or DataFrame)
• Check if title has HTML contents and filter them

We provide a colab notebook that helps you find target products and obtain their reviews!

We appreciate any help or feedback to improve the quality of our dataset! Feel free to reach us at jin018@ucsd.edu if you meet any following questions:

• Unparsed HTML contents
• Duplicate items which have same reviews

Directory

Files

Complete review data

Please only download these (large!) files if you really need them. We recommend using the smaller datasets (i.e. k-core and CSV files) as shown in the next section.

raw review data (34gb) - all 233.1 million reviews

ratings only (6.7gb) - same as above, in csv form without reviews or metadata

5-core (14.3gb) - subset of the data in which all users and items have at least 5 reviews (75.26 million reviews)

Per-category data - the review and product metadata for each category.

To download the complete review data and the per-category files, the following links will direct you to enter a form. Please contact me if you can't get access to the form.

"Small" subsets for experimentation

If you're using this data for a class project (or similar) please consider using one of these smaller datasets below before requesting the larger files.

K-cores (i.e., dense subsets): These data have been reduced to extract the k-core, such that each of the remaining users and items have k reviews each.

Ratings only: These datasets include no metadata or reviews, but only (item,user,rating,timestamp) tuples. Thus they are suitable for use with mymedialite (or similar) packages.

You can directly download the following smaller per-category datasets.

Data format

Format is one-review-per-line in json. See examples below for further help reading the data.

Sample review:

where

• reviewerID - ID of the reviewer, e.g. A2SUAM1J3GNN3B
• asin - ID of the product, e.g. 0000013714
• reviewerName - name of the reviewer
• vote - helpful votes of the review
• style - a disctionary of the product metadata, e.g., "Format" is "Hardcover"
• reviewText - text of the review
• overall - rating of the product
• summary - summary of the review
• unixReviewTime - time of the review (unix time)
• reviewTime - time of the review (raw)
• image - images that users post after they have received the product

Metadata

Metadata includes descriptions, price, sales-rank, brand info, and co-purchasing links:

metadata (24gb) - metadata for 15.5 million products

Sample metadata:

where

• asin - ID of the product, e.g. 0000031852
• title - name of the product
• feature - bullet-point format features of the product
• description - description of the product
• price - price in US dollars (at time of crawl)
• imageURL - url of the product image
• imageURL - url of the high resolution product image
• related - related products (also bought, also viewed, bought together, buy after viewing)
• salesRank - sales rank information
• brand - brand name
• categories - list of categories the product belongs to
• tech1 - the first technical detail table of the product
• tech2 - the second technical detail table of the product
• similar - similar product table

Code

Reading the data

Data can be treated as python dictionary objects. A simple script to read any of the above the data is as follows:

Pandas data frame

This code reads the data into a pandas data frame:

Example: compute average rating

Example: latent-factor model in mymedialite

Predicts ratings from a rating-only CSV file