Amazon Product Reviews Dataset

Dataset Overview

Data Type	Real	Default Task	Classification,Regression
Attribute Type	Multivariate, Sequential, Temporal aspect	Published Year	2018
Area of Dataset	Includes ratings, review text, and helpfulness votes	Missing Values	Yes
No. of Instances	233 million	No. of Attribute	16+

Dataset Description:

This dataset is an updated and expanded version of the Amazon review dataset originally released in 2014. It contains a comprehensive collection of customer reviews, product metadata, and relational links useful for recommendation systems and data analysis.

Key Features:

Large Scale: Over 233 million reviews spanning from May 1996 to October 2018 (compared to 142.8 million in the 2014 release).
Review Data: Includes ratings, review text, and helpfulness votes.
Rich Metadata: Enhanced product information such as color, size, package type, bullet-point descriptions, technical details (attribute-value pairs), and images taken post-purchase.
Relational Links: Graph data showing “also viewed” and “also bought” product relationships.
Expanded Categories: Five new product categories added to cover a wider range of items.

Applications:

Ideal for building and evaluating machine learning models in sentiment analysis, recommendation systems, opinion mining, and other NLP and data mining tasks.