Public release of Haitian Creole text data by Carnegie Mellon

See the main Haitian Creole page here for an overview of all available Haitian Creole speech and text data. This page describes the text data only.

Haitian Creole text data

Data license

• Medical domain phrases and sentences in English collected at Carnegie Mellon under the NSF-funded (jointly with the EU) NESPOLE! project, translated into Haitian Creole by Eriksen Translations, Inc.
• Zipped RAR file, English and Creole sentences together (31 KB)
• Plain text version, English and Creole sentences together (101 KB)
• English (50 KB) and Creole (49 KB) parallel files, one sentence per line
• English term index in PDF format (128 KB) mapping base-form nouns and verbs to the sentence numbers they appear in, and the corresponding numbered list of sentences in PDF format (377 KB)
• Newswire data collected at Carnegie Mellon under the DARPA-funded DIPLOMAT project
• English (2 MB) and Creole (2 MB) parallel files, one sentence per line
• Short phrases and other glossary entries collected and produced at Carnegie Mellon under the DIPLOMAT project
• English (519 KB) and Creole (499 KB) parallel files, one sentence per line.
  Update: Several people noticed a large amount of Creole text in the English file, and vice versa. This error has been corrected as of 1 p.m. EST on Wednesday, January 27, 2010. Please download these updated versions if required.

Contact for this page: Robert Frederking.
Last updated: Wednesday, January 27, 2010, 12:55 p.m. EST by Greg Hanneman.