Challenges in Annotating and Parsing Spoken, Code-switched, Frisian-Dutch Data

Keyphrases

• Best Model 33%
• Biaffine 33%
• Code-switching 100%
• Data Selection 33%
• Diacritics 33%
• Dutch 100%
• High Performance 33%
• Lag behind 33%
• Language Performance 33%
• Latent Dirichlet Allocation 33%
• Low-resource Languages 66%
• MBERT 33%
• N-gram 33%
• Parser 66%
• Parsing 100%
• Spoken Data 33%
• Target Domain 33%
• Training Data 33%
• Treebank 100%
• XLM-R 33%

Computer Science

• Language Resource 100%
• Latent Dirichlet Allocation 33%
• Parsing 100%
• Training Data 33%