Implementation of Phonetic Algorithm for Spelling Correction in Bahasa Indonesia

Abstract

A technique to obtain documents containing information in accordance with a user’s requirements is necessary for an information retrieval system. This research compared the performance of four phonetic algorithms: Soundex, Phonix, Metaphone, and Caverphone. This research used a database derived from KBBI 2005. The accuracy of Soundex, Phonix, Metaphone, and Caverphone is 92%, 78%, 77%, and 75%, respectively. However, Soundex and Phonix suggested twice as many words as Metaphone and Caverphone