Amazon Reviews Multi: Multilingual Customer Feedback Corpus

Multilingual Amazon customer reviews hosted as raw JSONL.GZ files for direct loading. The dataset is a mirror of the original 'amazon_reviews_multi' corpus, uploaded by user goosmanlei. It was last updated on the platform on 2026-03-18.

Use Cases

• Train multilingual sentiment analysis models based on the review text.
• Benchmark machine translation systems using parallel review content across languages.
• Conduct cross-lingual text classification research based on the multilingual corpus.
• Analyze e-commerce customer feedback patterns across different regions and languages.

Strengths

• Data is stored in a compressed JSONL.GZ format optimized for direct loading via the Hugging Face datasets library.
• The dataset is a mirror of a known multilingual corpus, suggesting a structured and potentially validated source.

Limitations

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

Provenance

Source

Mirror of the original 'amazon_reviews_multi' dataset from Hugging Face.

Collection Method

Raw JSONL.GZ data files retrieved from a user mirror.

Time Range

null

Freshness

Last updated 2026-03-18 17:16:24; freshness should be verified.

Geography

null