Amazon random products images and data schema

Ecommerce dataset from amazon

This dataset contains a large-scale collection of Amazon random products with images and structured metadata, designed for ecommerce research, machine learning, and application development. Each product record is linked to locally stored product images, ensuring fast access without external dependencies.

All product images have been downloaded and stored locally, and the corresponding local image file paths are attached to each product record. This makes the dataset ideal for offline processing, computer vision tasks, product classification, recommendation systems, and UI prototyping.

The dataset includes over 99,000+ unique product records and more than 493,000+ high-quality product images, with multiple images available per product. Metadata is organized in a clean, structured schema that supports seamless integration with databases, APIs, and machine learning pipelines.

This dataset can be used for:

• Product image recognition and similarity search

• Ecommerce recommendation engines

• Data science and analytics projects

• AI/ML model training and testing

• Marketplace UI/UX demos and mock stores

Data points:

url, asin, name, price, images, brand, product_overview, about_item, specifications, product_details, primary_category, category_1, category_2, category_3, breadcrumbs, availability, description, five_star, four_star, three_star, two_star, one_star, uniqId, scrapedAt, image_file