STEMMING IN MADURESE LANGUAGE USING NAZIEF AND ADRIANI ALGORITHM

Madurese is one of the regional languages in Indonesia, which dominates East Java and Madura Island in particular. However, the use of Madurese is declining compared to other regional languages. This is partly due to a sense of prestige and difficulty in learning it. As a result, the future of Madurese as one of the regional languages in Indonesia is increasingly threatened by the decline in its use. In addition, academic literature and scientific publications in Madurese are difficult to find in public and academic libraries, so previous research on Madurese stemming is still very little and needs to be developed further. Therefore, this research aims to find the base word of Madurese language using Nazief & Adriani algorithm based on Madurese language morphology. The Nazief & Adriani method in previous studies has good performance. Stemming can also be developed into a Madurese language translator application into other languages. This research uses 650 words in the form of datasets, consisting of 500 prefix words and 150 suffix words. The resulting accuracy for the whole is 96.61% with 628 correct words, the prefix has 95.6% accuracy, and the suffix has 100% accuracy. Overstemming was found in 22 prefix words and no words experienced Understemming.