Shopping Queries Image Dataset: 190,000+ Products with Images for Multimodal Search

An augmented version of the Amazon Shopping Queries Dataset, SQID includes image information for over 190,000 products. The dataset was created by crossingminds and last updated on Hugging Face in September 2024. It pairs product search queries from real Amazon users with up to 40 potentially relevant results.

Use Cases

• Train multimodal product search models based on the combination of text queries and product images.
• Benchmark image-text retrieval systems based on the dataset's query-result pairs.
• Explore relevance ranking for e-commerce based on the list of up to 40 results per query.
• Fine-tune vision-language models for product understanding based on the image-enriched product data.

Strengths

• Includes image information for over 190,000 products, providing a substantial visual component.
• Augments a known benchmark (Amazon Shopping Queries Dataset) with multimodal data.
• Based on real user search queries from Amazon, suggesting practical relevance.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

crossingminds on Hugging Face, augmenting the Amazon Shopping Queries Dataset.

Collection Method

Augmentation of an existing dataset with image information; specific collection method not detailed.

Time Range

null

Freshness

Last updated 2024-09-04 20:38:39.

Geography

Regionus tag suggests a US focus, but not explicitly confirmed.