Amazon-Baby Datasets
Dataset Description
The Amazon-Baby dataset is a shopping network. It includes information about books, users and reviews. Nodes represent items and users. Text on a item node is the description of the item. Text on a user node is the reviwer
. The item text includes the following information: This product, available from [store] and called [title], has an average rating of [average_rating]. Key features include [features], and according to the description, [description]. The price is [price], and additional details [details].
Edges represent relationships between items and users. Text on an edge is a user's review of a item including following information: This review, titled [title], gives the product a rating of [rating]. The reviewer stated, [text]
.
Graph Machine Learning Tasks
Link Prediction
Link prediction in the Amazon-Baby dataset involves predicting potential connections between users and items. The goal is to predict whether a user will purchase a item.
Node Classification
Node classification tasks in the Amazon-Baby dataset include predicting the items's category.
Dataset Source
https://amazon-reviews-2023.github.io/data_processing/5core.html