smartcat/Amazon_Sample_Metadata_2023

Dataset Card for Dataset Name

Original datasets can be found on: https://amazon-reviews-2023.github.io/

Dataset Details

This dataset was made as sample of several datasets from the link above.

Dataset Description

This dataset is a curated sample derived from seven filtered Amazon product category datasets(Amazon All Beauty, Amazon Fashion, Sports and Outdoors, Health and Personal Care, Amazon Clothing Shoes and Jewlery, Baby Products and Beauty and Personal Care), combining diverse product information across categories. The dataset contains 258k entries, each representing a unique product and its associated metadata. Designed for product-related analysis, this dataset is well-suited for tasks such as embedding generation, recommendation systems, and NLP model training.

Dataset Versions:

• Data version: Contains standard product attributes, without the additional text column.
• Formatted version: Includes an extra column, text, specifically created for generating embeddings, allowing for deeper semantic analysis and machine learning tasks.
• Embeddings: Generated embeddings from the text column, along with parent_asin and token count.
• Formatted and embedded version: The formatted version with the embeddings.
• Image description version: The standard dataset with the description of the images for each product.
• Image embeddings: The standard dataset with the image embedding of the first encountered image of each product.
• Combined description version: The standard dataset along with the image description and combined description and image description in the image_description_formatted column.
• Combined and formatted description version: The standard dataset along with the image description, and combined description and image description in the image_description_formatted column. Including the text column.
• Embeddings for Combined description version: The Combined and formatted description version along with token_counts and embeddings of the text column.
• Query columns version: The standard dataset along with short queries and long queries.
• Product2Query version: The dataset with query columns, splited into train and test for fine-tuning.

Original Structure

Field	Type	Explanation
main_category	str	Main category (i.e., domain) of the product.
title	str	Name of the product.
average_rating	float	Rating of the product shown on the product page.
rating_number	int	Number of ratings in the product.
features	list	Bullet-point format features of the product.
description	list	Description of the product.
price	float	Price in US dollars (at time of crawling).
images	list	Images of the product. Each image has different sizes (thumb, large, hi_res). The “variant” field shows the position of image.
videos	list	Videos of the product including title and url.
store	str	Store name of the product.
categories	list	Hierarchical categories of the product.
details	dict	Product details, including materials, brand, sizes, etc.
parent_asin	str	Parent ID of the product.
bought_together	list	Recommended bundles from the websites.

Modifications made

• Products without a description, title, images or details were removed.
• Lists in features and description are transformed into strings concatinated with a newline
• For the details column, only the top 16 most frequent detail types were kept. The details column was then split into these new 16 columns based on the detail types kept.
• Products with date first available before the year 2015 are dropped.
• Products with is_discontinued_by_manufacturer set to 'true' or 'yes' are dropped. Then that column was dropped.
• Column bought_together is dropped due to missing values.

Dataset Size

• Total entries: 258k
• Total columns (Data version): 45

Final Structure

Field	Type	Explanation
main_category	str	Main category
title	str	Name of the product
average_rating	float	Rating of the product shown on the product page.
rating_number	int	Number of ratings in the product.
features	list	Bullet-point format features of the product.
description	list	Description of the product.
price	float	Price in US dollars (at time of crawling).
images	list	Images of the product. Each image has different sizes (thumb, large, hi_res). The “variant” field shows the position of image.
videos	list	Videos of the product including title and url.
store	str	Store name of the product.
details	dict	Product details, including materials, brand, sizes, etc.
parent_asin	str	Parent ID of the product.
upc	str	Universal Product Code (UPC), a barcode for uniquely identifying the product.
brand	str	Brand
manufacturer	str	Manufacturer of the product
product_dimensions	str	Dimensions of the product
color	str	Color
material	str	Material
item_form	str	Form of the item (e.g., liquid, solid, gel).
item_model_number	str	Model number of the product
age_range_(description)	str	Recommended age range for the product, often used for toys or children’s products.
item_weight	str	Weight of the item
skin_type	str	Suitable skin types for the product (e.g., oily, dry, sensitive).
package_dimensions	str	Dimensions of the product packaging
hair_type	str	Suitable hair types for the product (e.g., curly, straight, fine).
unit_count	str	Total quantity or units contained in the product package.
number_of_items	float64	Number of items in the product
scent	str	Fragrance or scent associated with the product.
material_type	str	Material
style	str	Style
date_first_available	float64	Date the first time product was available
best_sellers_rank	str	Best seller rank
department	str	Department
size	str	Sieze
country_of_origin	str	Country of origin of the product
flavor	str	Flavour
item_dimensions_lxwxh	str	Item dimensions LxWxH
brand_name	str	Brand name
package_weight	str	Weight of the package
item_package_dimensions_l_x_w_x_h	str	Dimensions of the item package LxWxH
part_number	str	Part number
suggested_users	str	Suggester users
item_dimensions__lxwxh	str	Item dimensions LxWxH
sport_type	str	Sport type
closure_type	str	Closure type