Amazigh-English Tatoeba Extended: Parallel Sentences with Synthetic Variations

Amazigh-English Tatoeba Extended provides original Standard Moroccan Amazigh sentences paired with English translations. The dataset is augmented with a multi-tiered set of 6 synthetic English variations per entry. It was created by abdelhaqueidali and last updated on June 4, 2026.

Use Cases

• Train machine translation models based on the Amazigh-English parallel sentences.
• Develop text alignment algorithms using the paired original and synthetic variations.
• Augment NLP training data for Amazigh language tasks using the 6 synthetic variations per entry.
• Benchmark translation quality for the Amazigh language.
• Study the effect of synthetic rephrasing on model performance.

Strengths

• Includes 6 synthetic English variations per original entry, providing multiple training examples.
• Original sentences are written in the Amazigh script/alphabet, supporting language preservation.

Limitations

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

Provenance

Source

Tatoeba, augmented by abdelhaqueidali.

Collection Method

Original Tatoeba parallel sentences augmented with synthetic English variations.

Freshness

Last updated 2026-06-04 12:46:24; freshness should be verified.