PASS: Automatically Generated Summaries for 32 Amazon Product Reviews

Amazon's PASS dataset contains automatically generated summaries for product reviews. The summaries were produced by the Perturb-and-Select Summarizer (PASS) method for 32 products sourced from the FewSum dataset. The dataset is hosted on AWS Open Data and was created by Amazon.

Use Cases

• Benchmarking automated summarization models based on the generated review summaries.
• Training or fine-tuning abstractive summarization systems using the review-summary pairs.
• Analyzing the quality and characteristics of machine-generated summaries for e-commerce text.

Strengths

• Summaries are generated for a specific, known source dataset (FewSum).
• Focuses on a defined domain of product reviews.
• Originates from a major research organization (Amazon).

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Last update date is unknown; freshness unverified.

Provenance

Source

Amazon

Collection Method

Summaries automatically generated by the PASS (Perturb-and-Select Summarizer) method.

Time Range

null

Freshness

Last update date is unknown; freshness unverified.

Geography

null