Abstract—an efficient stop-word removal technique is needed in many natural languages processing application such as: spelling normalization, stemming and stem weighting, Question, Answering systems and an Information Retrieval system (IRS). Most of the existing stop-word removal techniques are based on a dictionary that contains a list of stop-word which is very expensive and takes much time for searching process and required much space to store these stop-words. So, the main objective of this paper is to design and implement an efficient Information Retrieval System for Arabic language. The proposed technique is based on analyzing document in a statistical way through two approaches: manual methods for IRS and automatic methods for IRS. The new proposed Arabic removal stop-word technique has been tested using a set of 500 Arabic abstracts chosen from a set of data chosen from different resources and it gives impressive results. Our work presented in this paper is based on analyzing document in a statistical way through two approaches: (1) Manual method for (IRS) in which we should counting the number of repetition of the word in the document then comparing the degree of similarity between the counted repetition words through using some static database contain specific subject. (2) Automatic methods for (IRS) in which we should counting the number of repetition of the word in the document after that calculating the weight of the counted repetition words through using two different algorithms using mathematical formulas. 

Index Terms—automatic indexing, evaluation, information retrieval system, stop word.

Cite: Mohammad O.Wedyan and Aarti Singh, "On the Design and Implementation of an Efficient Information Retrieval System for Arabic Language," International Journal of Electronics and Electrical Engineering, , Vol. 1, No. 1, pp. 5-9, March 2013. doi: 10.12720/ijeee.1.1.5-9