A Combined Spell Checking and Error Correcting System for Punjabi -Hindi Language using Hybrid Approach
Amandeep Singh · International Journal of Advanced Research in Computer Science · 2016
Spellchecker is software that analyzes possible misspellings in the text. It is the process of detecting and sometimes providing some suggestions for incorrectly spelled words in a text. If dictionary of spell checker is larger than higher is the error detection and error correction rate. Though considerable work has been done in English language but not much work has been done in regional languages of India including Punjabi and Hindi. Punjabi is the official language of Punjab state in India Punjabi is world’s 12th most widely spoken language In Punjabi language, there is very small amount of work is completed in this region.. Hindi is the official language of 11 vowels and 33 consonants. Hindi is also the third most spoken language in the world. A few work is done in Punjabi-Hindi spell detection and correction field and it is not easy task to identify errors in Punjabi-Hindi text. The spell checker systems are online available but as not stand-alone applications. The only available spell checker for Punjabi is “Akhar” ”Raftaar” and “Sudhaar”. “Akhar” is paid software that is not available free for its use to everybody and “Sudhaar” spell checker is a desktop application Some paid Hindi spell checker software’s are also online available. “Hinspell” & “Hinkhoj” are available spell checkers for Hindi language but a lot of improvement is needed. NLP (Natural language processing) is a field of computer science concerned with interaction between computer and human language. We have developed a combined spell checker and error correcting system for both Punjabi and Hindi Language. We used hybrid approach to implement the Spelling checking and Correcting System. The proposed system use hybrid approach which s a combination of various approaches like rule based approach, dictionary lookup approach , edit distance approach and NGram approach. Proposed system is tested with various inputs collected from different sources and results are found very accurate than that of existing system. Keywords: Spell Checking, Error Detection, Error Correction, Punjabi, Hindi, Dictionary-lookup, Hybrid approach.