568,454 consumer reviews across various Amazon product categories. Records consist of textual feedback, star ratings, and user-product identifiers.
Use Cases
- Train a sentiment classifier to predict the 'Score' using the 'Text' column.
- Build a collaborative filtering model using the 'UserId' and 'ProductId' columns.
- Analyze the helpfulness of reviews by calculating the ratio between 'HelpfulnessNumerator' and 'HelpfulnessDenominator'.
Strengths
- 568,454 rows of review data.
- Includes 'Score' column with integer values from 1 to 5.
- Contains 'Text' and 'Summary' string fields for NLP.
- Features 'UserId' and 'ProductId' alphanumeric identifiers.
- Includes 'Time' column in Unix timestamp format.