Review of Real-word Error Detection and Correction Methods in Text Documents

Abstract

Spell checking is one of the most widely used tasks of NLP. It has broad range of uses like information retrieval, proofreading, email client etc. Today in many applications of NLP spell checker is being used. It is a language tool which breaks down the text for spelling errors. It flags when there exists any misspelled/unintended word in the given text. The typographic errors are categorized in ‘non-word errors’ and ‘real-word errors’. There is enough work done in order to tackle the farmer error but it still remains the challenge for the researchers to tackle the later one. This paper focuses on later one and analyses the methods which are being used worldwide to detect and correct such errors. Paper also focuses on the challenges faced by researchers while processing the real-word errors.