Developing a spell checker

Abstract

Spell checkers are ubiquitous computational tools that help us in correctly writing texts or messages and improving information inquiry and data mining. The present work presents the history of development of spell checkers and illustrates how, in a simple way, it is possible to create an efficient spell checker from Norvig’s proposal. We also highlight some tools and how they are used in the development of spell checkers, such as affix removal and n-gram computation. Moreover, we present an implementation of Norvig’s spell checker and its performance in automatic correction for different spelling error data sets. Also, in a comparison of spell checkers performance, we expose that it is worth removing affixes.

Keywords:
Spell checker; Spelling; Orthography; Affixes; Computational linguistics