Stemming Javanese: Another Adaptation of the Nazief-Adriani Algorithm

Abstract

Javanese is the regional language with the most speakers in Indonesia. Not only in the Java island, but the Javanese is also spoken by people outside Indonesia, such as in Malaysia and Suriname. Besides Bahasa Indonesia, which is the national language that must be learned by Indonesian society, the regional languages, like Javanese, also must be preserved. One of the preservation methods is by using an information retrieval system. One of the popular preprocessing methods in information retrieval is called stemming. It is used to reduce differences in word forms so that the information retrieval process becomes effective. Since Javanese has its uniqueness, the language morphology is different from Bahasa Indonesia. Thus, the stemming process of Javanese words also differs from Bahasa Indonesia. The stemming algorithm in this paper is developed by adapting the Nazief-Adriani algorithm, a well-known Bahasa Indonesia stemmer method. The algorithm in this paper is made based on the rules of Javanese language morphology. It removes ater-ater (prefixes), seselan (infixes), penambang (suffixes), bebarengan (confixes), and tembung rangkep (repeated word).