konect logo
KONECT
KONECT > Networks > Amazon ratings

Amazon ratings

About this network

This bipartite network contains product ratings from the Amazon online shopping website. The rating scale ranges from 1 to 5, where 5 denotes the most positive rating. Nodes represent users and products, and edges represent individual ratings.

Network info

CodeAR
Category Rating
Data source http://liu.cs.uic.edu/download/data/
Vertex type User, product
Edge type Rate
FormatBipartite: Edges connect two types of nodes Bipartite
Edge weightsMultiple ratings: Multiple weighted edges are allowed Multiple ratings
Metadata Timestamps:  Edges are annotated with a timestamps TimestampsIncomplete:   not all edges or nodes from the original dataset are included Incomplete
Size5,523,029 = 3,376,972 + 2,146,057 vertices (users + products)
Volume5,838,041 edges (rates)
Unique volume5,743,258 edges (rates)
Average degree (overall)3.4576 edges / vertex
Average user degree2.7204 edges / vertex
Average product degree4.7428 edges / vertex
Fill2.1741 10–6 edges / vertex2
Maximum degree12,217 edges
Negativity37.7%
Wedge count627,186,651
Claw count704,564,362,291
Square count35,849,304
4-tour count2,807,043,116
Power law exponent (estimated) with dmin2.0710 (dmin = 1)
Gini coefficient65.1%
Relative edge distribution entropy91.7%
Assortativity–0.035658
Diameter28 edges
90-percentile effective diameter8.04 edges
Mean shortest path length6.63 edges
Spectral norm745.70
Preferential attachment exponent0.60480 (ε = 2.5231)
Temporal distribution of the Amazon ratings network
Temporal distribution
Edge weight distribution of the Amazon ratings network
Edge weight distribution
Edge multiplicity distribution of the Amazon ratings network
Edge multiplicity distribution
Cumulative edge multiplicity distribution of the Amazon ratings network
Cumulative edge multiplicity distribution
User degree distribution of the Amazon ratings network
User degree distribution
Product degree distribution of the Amazon ratings network
Product degree distribution
User degree distribution of the Amazon ratings network
User degree distribution
Product degree distribution of the Amazon ratings network
Product degree distribution
Degree distribution of the Amazon ratings network
Degree distribution
User degree distribution of the Amazon ratings network
User degree distribution
Product degree distribution of the Amazon ratings network
Product degree distribution
Distance distribution of the Amazon ratings network
Distance distribution
Distance distribution on a logistic scale of the Amazon ratings network
Distance distribution on a logistic scale
Spectral distribution of the eigenvalues of A of the Amazon ratings network
Spectral distribution of the eigenvalues of A
Spectral distribution of the eigenvalues of N of the Amazon ratings network
Spectral distribution of the eigenvalues of N
Spectral distribution of the eigenvalues of L of the Amazon ratings network
Spectral distribution of the eigenvalues of L
Cumulative spectral distribution of A of the Amazon ratings network
Cumulative spectral distribution of A
Cumulative spectral distribution of N of the Amazon ratings network
Cumulative spectral distribution of N
Cumulative spectral distribution of L of the Amazon ratings network
Cumulative spectral distribution of L

Downloads

TSV file:downloadamazon-ratings.tar.bz2 (37.73 MiB)
Extraction code:downloadamazon-ratings.tar.bz2 (14.51 KiB)

References

[1] Amazon ratings network dataset -- KONECT, April 2017. [ http ]
[2] Ee-Peng Lim, Viet-An Nguyen, Nitin Jindal, Bing Liu, and Hady Wirawan Lauw. Detecting product review spammers using rating behaviors. In Proc. Int. Conf. on Information and Knowledge Management, pages 939--948, 2010.
[3] Nitin Jindal and Bing Liu. Opinion spam and analysis. In Proc. Int. Conf. on Web Search and Web Data Mining, pages 219--230, 2008.
[4] Arjun Mukherjee, Bing Liu, and Natalie Glance. Spotting fake reviewer groups in consumer reviews. In Proc. Int. Conf. on World Wide Web, pages 191--200, 2012.

BibTeX