An Efficient Autocorrect Algorithm Using Genetic Algorithm

Sol Kim
Farragut High School

Abstract

Autocorrect is used widely in digital typing to, as the name suggests, automatically correct mistyped words. The purpose of this paper is to investigate what factors should contribute to how a mistyped word is corrected and how significant each factor is in the correction process. First, the aspects that could contribute to a more efficient autocorrect algorithm include how different a mistyped word is from the intended word, how far away a mistyped character is from the correct one, and how the typo is classified. These factors were then quantified and integrated into the autocorrect algorithm to see how accurately a corpus of words could be corrected. Finally, these quantities that were assigned were altered and experimented with to inspect how much they contributed to the accuracy of the algorithm.

Please click to read the paper.