ARABIC LIGHT STEMMER (ARS)

Stemming is a main step used to process textual data. It is usually used in several types of applications such as: text mining, information retrieval (IR), and natural language processing (NLP). A major task in stemming is to standardize words; which can be achieved by reducing each word to its base (root or stem). Arabic stemming is not an easy task. Unlike other languages, Arabic language is a highly inflected language, since it uses many inflectional forms. Researchers are divided on the benefit of using stemming in fields of IR, NLP... etc., since in Arabic the morphological variants of a certain word are not always semantically related. The aim of this paper is to design and implement a new Arabic light stemmer (ARS) which is not based on Arabic root patterns. Instead, it depends on well defined mathematical rules and several relations between letters. A series of tests were conducted on ARS stemmer to compare its effectiveness with the effectiveness of two other Arabic stemmers. Test shows clearly the effectiveness superiority of ARS compared to effectiveness of these two Arabic stemmers.