Papers with misspelled term identification
English Language Spelling Correction as an Information Retrieval Task Using Wikipedia Search Statistics (2022.lrec-1)
Copied to clipboard
| Challenge: | Existing spelling correction tools lack training or annotated data sets to perform . many spelling correction utilities suffer due to the size and quality of dictionaries available to aid correction. |
| Approach: | They propose a dynamic spelling correction tool that uses the Wikipedia dataset search API to aid misspelled term identification and automatic replacement. |
| Outcome: | The proposed spelling correction tool performs comparable to the industry-standard spelling correction algorithm, Hunspell. |