amazon_reviews — Amazon product reviews (1996-2018)

Description

Bipartite networks of reviewers and products derived from the Amazon review data collected by J. Ni, J. Li and J. McAuley (source), which contains product reviews (with ratings, text and product metadata) from the Amazon online store, spanning May 1996 to October 2018. The networks were assembled by A. R. Benson and collaborators. Reviewers and products are anonymized and given only as integer indices.

reviews: products and the users who reviewed them, built from the "5-core" subsets of the data, in which, for each of the 29 product categories, only the users and products with at least five reviews in that category were kept. Each user is connected to every product they reviewed. The 33 products listed in the source data without any reviewer were removed.
music_blues: reviewers of music products tagged "regional blues" (a subset of vinyl records), grouped by time and sub-genre. The reviews were split into calendar months, and for each month and each sub-genre (e.g. "Chicago Blues") there is a group node connected to all users who reviewed a product of that sub-genre during that month. The same sub-genre can therefore give rise to many groups, and identical sets of users can appear in several groups. The reviews span the years 1998 to 2018, but the month of each group is not given in the source data. Users who belong to no group were removed.

The networks have the following properties:

reviews: is_user, whether the node is a user (otherwise a product); category, the product category (e.g. "Appliances", "Video_Games"; empty for users).
music_blues: is_group, whether the node is a group of users who reviewed products of a given sub-genre in a given month (otherwise a user); genre, the blues sub-genre of the group (empty for users).

Tags

Economic Commerce Online Preferences Unweighted Metadata

Citation

J. Ni, J. Li, J. McAuley, "Justifying Recommendations using Distantly-Labeled Reviews and Fine-Grained Aspects." Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), 188-197 (2019) 10.18653/v1/D19-1018 Sci-Hub
N. Veldt, A. R. Benson, J. Kleinberg, "Minimizing Localized Ratio Cut Objectives in Hypergraphs." Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 1708-1718 (2020) 10.1145/3394486.3403222 Sci-Hub
I. Amburg, N. Veldt, A. R. Benson, "Diverse and Experienced Group Discovery via Hypergraph Clustering." Proceedings of the 2022 SIAM International Conference on Data Mining (SDM), 145-153 (2022) 10.1137/1.9781611977172.17 Sci-Hub

Upstream URL OK

https://www.cs.cornell.edu/~arb/data/amazon-reviews/
https://www.cs.cornell.edu/~arb/data/cat-edge-music-blues-reviews/

Networks

Tip: click on a column header to sort the list, or hover over it for a description.

Column legend

$\left<k\right>$: Mean degree
$\sigma_k$: Standard deviation of the degrees
$\lambda_h$: Largest eigenvalue of the Hashimoto (non-backtracking) matrix
$\tau$: Random walk mixing time
$r$: Degree assortativity
$c$: Global clustering coefficient
$\oslash$: (Pseudo-) diameter
$S$: Relative size of the largest component
NPs: Node properties
EPs: Edge properties

Name	Nodes	Edges	$\left<k\right>$	$\sigma_k$	$\lambda_h$	$\tau$	$r$	$c$	$\oslash$	$S$	Kind	Mode	NPs	EPs	gt	GraphML	GML	csv
reviews	6,553,594	73,141,425	22.32	78.49	401.76	118.57	-0.05	0.00	11	1.00	Undirected	Bipartite	is_user category		217.2 MiB	319.3 MiB	298.6 MiB	348.3 MiB
music_blues	1,800	10,499	11.67	12.71	24.91	104.82	-0.16	0.00	16	1.00	Undirected	Bipartite	is_group genre		14 KiB	32 KiB	31 KiB	34 KiB