Manually Annotated Spelling Error Corpus for Amharic

This addresses a problem for NLP researchers and developers working on Amharic language processing by providing a foundational dataset, though it is incremental as it builds on existing corpus creation methods.

The paper tackles the lack of resources for spelling error detection and correction in Amharic by creating a manually annotated corpus with 1,000 sentences and 5,000 tokens, tagged for non-word and real-word errors, which enables evaluation and handling of both error types.