Skip to Main Content
This paper examines the performance of Addaall stemmers for Arabic news of al-Jazeera. The morphological variation of Arabic word is a factor that affects Arabic information retrieval performance. Many morphological stemmers were developed to improve Arabic information retrieval such as Buckwalter, Khoja stemmers. Addaall stemmer is among the recent developments pertaining to the enhancement of the effectiveness of Arabic IR. The objective of the paper is to provide an analysis of Addaall's stemmer performance for Arabic information. Furthermore, it sheds some lights on the effectiveness of Addaall in relation to other stemmers. The paper assumes that previous studies did not examine extensively the requirements and limitations of Addaall and its prospects for Arabic IR effectiveness. So far there is no standard stemmer for Arabic IR that can be used and generalized. Therefore, this research will examine and evaluate the retrieval performance of Addaall stemmer using root and light stemming searches. It will use Al-Jazeera news as a test document. The paper concludes that Addaall offers a horizon for further enhancing the effectiveness of Arabic IR.