Clustering the Arabic documents (CAD) Boumedyen Shannaq Nizwa university Abstract We present new clustering techniques to discover when two documents have the same content; the proposed techniques yields a grouping of the documents into clusters of closely related items, this clustering can help solve the problems of document duplication. The new clustering techniques based on LIPNS techniques. We believe that our system provides new functionality for dealing with huge amount of information on the web .It allows search engines to better present results to their clients and to measure the similarity of any two objects or cluster the sets of similar documents from a large corpus. The new clustering technique was implemented on Arabic documents.