Dataset Overview

The Amazon Review Data (2018) is an extensive collection of customer reviews and product metadata spanning from May 1996 to October 2018. It serves as a major benchmark for research in recommendation systems, natural language processing, and consumer behavior analysis. The full dataset contains approximately 233.1 million reviews across dozens of product categories, including detailed information such as ratings, text, product images, and co-purchasing links. To facilitate experimentation, the data is available in various scales, ranging from the massive raw review files to smaller "5-core" subsets where each user and item has at least five reviews. More

Total node count: 323,139,730

Total edge count: 742,265,681

Dataset Download

You may download this dataset directly from the dataset website. Go to here, scroll down to the File section. To download the complete dataset, you may be required to complete a Google form provided by the dataset authors.

You can click the button below and download the files prepared by us for smoother import into graph database systems:

Benchmark Workloads

Select a platform below to view the specific data loading commands and queries used to benchmark performance.

Runtime Results

* The unit of all results shown below are in milliseconds (ms)