Sistem Deteksi Struktur Kalimat Bahasa Arab Menggunakan Algoritma Light Stemming

Abstract

To understand Arabic, it is necessary to study nahwu theory. Nahwu theory is the study of structure of Arabic sentences. By learning nahwu, being able to distinguish subjects, predicates and objects in sentences. One of the fields in the computer that studies about human language processing is NLP (Natural Language Processing), which is natural human language processing through syntactic analysis of sentences structure. One method to analyzing syntactic of sentences is stemming. One variant of the stemming algorithm is Light Stemming algorithm. Light Stemming algorithm is an algorithm that only removes prefix and sufix. Based on the tests conducted, the light stemming algorithm is able to detect nahwu with an accuracy rate of 82.22%. The 17.78% failure rate occurs because words that do not have an affix will be detected automatically the type is fi’il (verb), even though the fact may be that the type is isim (noun), and the failure of the detection results is also due to not being able to stemming the words that have affix in the middle (infix), because indeed the process of light stemming only eliminates the sufix and prefix.