Introduces CHALIS benchmark dataset testing language ID on mutually intelligible cousin language pairs and orthographically noisy inputs, with evaluation showing existing systems struggle substantially.
Hierarchical Character- Models for Language Identification
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
CHALIS: A Challenge Dataset for Language Identification in Difficult Scenarios
Introduces CHALIS benchmark dataset testing language ID on mutually intelligible cousin language pairs and orthographically noisy inputs, with evaluation showing existing systems struggle substantially.