Ecommerce Product Reviews With Spanish Language Drift

Featuring product reviews from an e-commerce store, labeled with star ratings from 1 to 5. The training and validation sets consist of English reviews, while the production set introduces reviews in Spanish to simulate language drift.

Use Cases

• Train a sentiment classification model on English review text using the 1-5 star rating as a label.
• Detect and analyze language drift by monitoring model performance between the English training set and the Spanish-inclusive production set.
• Benchmark multilingual or cross-lingual text classifiers on the mixed-language production set.

Strengths

• Includes a controlled language drift scenario with Spanish reviews in the production set.
• Reviews are labeled on a 1-5 star scale for supervised learning tasks.
• Dataset is associated with a tutorial for practical application.

Limitations

• The specific number of rows, columns, and total size are unknown.
• The dataset's geographic and temporal coverage are not specified.
• The composition and balance of Spanish vs. English reviews in the production set are unclear.

Provenance

Source

Arize AI, derived from an extended IMDb source dataset.

Collection Method

Expert-generated annotations and language.

Freshness

Last updated on 2024-09-10.

Geography

Region is listed as US, but specific spatial coverage is unknown.