Development of Spell Corrector Model for Phrases
Rahul Singh Chauhan, Tanu Negi, Chandradeep Bhatt · 2023
This paper presents a spelling correction system integrated into an intelligent tutor, facilitating natural language dialogue. The system combines rule-based and statistical approaches to identify and correct spelling errors. By considering word frequency and contextual information, it estimates the probability of a word being misspelled. It employs an efficient method for correcting misspelled words, utilizing additions, deletions, and substitutions, ranked by statistical models. To enhance robustness, the system employs adaptive learning, regularly updating its dictionary and models based on user input. Extensive testing across various domains confirms its accuracy, recall, and improved typing efficiency.