Evaluation Campaign
>> Test Data

Test Sets for Translation of Manual Transcription

Contents

Every data package contains 1 test set. Languages other than English are segmented based on the ASR output.

Data Packages

• Test set(text) Chinese → English (Testset-Text-CE.tar.gz)
  
  
• Test set(text) English → Chinese (Testset-Text-EC.tar.gz)
  
  
• Test set(text) Japanese → English (Testset-Text-JE.tar.gz)
  
  
• Test set(text) Korean → English (Testset-Text-KE.tar.gz)
  
  
• Test set(text) Arabic → English (Testset-Text-AE.tar.gz)

Test Sets for Translation of Speech Recognizer output

Contents

Every data package contains 1 test set. Languages other than English are segmented based on the ASR output.

Data Packages

• Test set(lattices & n-best lists) Chinese → English
  (Testset-lattice-nbest-CE.tar.gz)
  
  
• Test set(lattices & n-best lists) English → Chinese
  (Testset-lattice-nbest-EC.tar.gz)
  
  
• Test set(lattices & n-best lists) Japanese → English
  (Testset-lattice-nbest-JE.tar.gz)