The Ngoko Javanese Stemmer uses the Enhanced Confix Stripping Stemmer Method

Abstract

Stemming is vital in text processing. The stemming that is most often encountered is Indonesian and English stemming. This is because more articles are processed in text processing in English and Indonesian. Indonesia has several regional languages, especially local school content, often used in learning. Therefore, research is needed to process Javanese language texts to make it easier for education practitioners, especially in Ngoko Javanese. Ngoko Javanese stemming, which still uses the affix removal stemmers method (rule-based approach) in previous research. Has a problem, namely the lack of success of this method when returning the root words of Ngoko Javanese, so it is necessary to check the Ngoko Javanese dictionary so that the results of the root words obtained are maximized. This study aims to conduct stemmer research on Ngoko Javanese using the Enhanced Confix Stripping (ECS) method. This stemmer is designed to do word splitting according to the Enhanced Confix Stripping algorithm and through checking the dictionary adapted to the Ngoko Javanese language. The results of this study are the ability to extract essential words in Javanese Ngoko to achieve a level of truth in returning root words reaching 97 percent.