DATA SOURCES AND COMMUNITY EVIDENCE
Where the data comes from.
Original data publishers and the research built on their records. Shared community datasets are planned for a later release.
SOURCE ARCHIVE · MCAULEY LAB
Amazon Reviews 2023
Collected and published by McAuley Lab at UC San Diego. The release contains Amazon reviews from multiple years. Gijs Overgoor and coauthors used reviews dated January 1–September 14, 2023 across all 34 release categories, including Unknown, for the reported audit.
The findings, thresholds, and category summaries shown in Public Traces come from the paper’s authors’ analysis of this archive.
OUR EARLIER PNAS RESEARCH · 2019–2020 DATA
Fake-reviews dataset and documentation
This data comes from our research published in PNAS in 2022: Detecting Fake Review Buyers Using Network Structure: Direct Evidence from Amazon.
Shared through coauthor Brett Hollenbeck’s repository, the public release also includes labels and variables from subsequent studies. The documentation explains those additions. This is a separate dataset from the McAuley Lab archive above.
The worked demos use generated records to explain the method and contribution process.
PLANNED NEXT STEP
A shared evidence base.
This release lets you explore the research and collect cases locally. It does not accept uploads or publish community datasets. A later contribution workflow is intended to retain sources, collection limits, and independent review before evidence becomes eligible for release.
Interested in using this for research or an investigation? Contact Gijs →