Kreyol-MT: Parallel Text for Creole Language Machine Translation

Kreyol-MT is a machine translation dataset for Creole languages, hosted by the Johns Hopkins University Center for Language and Speech Processing. The dataset is intended for release but is currently incomplete, pending a portion from the Linguistic Data Consortium. The repository was last updated on October 24, 2024.

Use Cases

• Train machine translation models based on parallel text for language pairs like 'acf-eng'.
• Benchmark translation quality for Creole languages based on the described parallel corpus.
• Conduct linguistic analysis of Creole language structures based on the translation pairs.
• Fine-tune multilingual language models on Creole language data based on the available language pairs.

Strengths

• Hosted by a reputable academic institution, Johns Hopkins University CLSP.
• Repository was updated on October 24, 2024, indicating recent maintenance.

Limitations

• The full dataset is not yet released; completeness is uncertain.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count, file formats, and license information are unknown, limiting suitability assessment.

Provenance

Source

Johns Hopkins University Center for Language and Speech Processing (jhu-clsp), with a portion from the Linguistic Data Consortium.

Collection Method

Likely compiled from existing translation resources or created for research.

Time Range

null

Freshness

Last updated 2024-10-24 18:31:46; freshness should be verified.

Geography

null