An Interactive Rough Set Attribute Reduction using Great Deluge Algorithm Najmeh Sadat Jaddi and Salwani Abdullah Data Mining and Optimization Research Group (DMO), Center for Artificial Intelligence Technology, Faculty of Information Science and Technology, Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor, Malaysia najmehjaddi@gmail.com,salwani@ftsm.ukm.my Abstract. Dimensionality reduction from an information system is a problem of eliminating unimportant attributes from the original set of attributes while avoiding loss of information in data mining process. In this process, a subset of attributes that is highly correlated with decision attributes is selected. In this paper, performance of the great deluge algorithm for rough set attribute reduction is investigated by comparing the method with other available approaches in the literature in terms of cardinality of obtained reducts (subsets), time required to obtain reducts, number of calculating dependency degree functions, number of rules generated by reducts, and the accuracy of the classification. An interactive interface is initially developed that user can easily select the parameters for reduction. This user interface is developed toward visual data mining. The carried out model has been tested on the standard datasets available in the UCI machine learning repository. Experimental results show the effectiveness of the method especially with relation to the time and accuracy of the classification using generated rules. The method outperformed other approaches in M-of-N, Exactly, and LED datasets with achieving 100% accuracy. Keywords: Interactive data mining, Great deluge algorithm, Rough set theory, Attribute reduction, Classification. 1 Introduction Visual data mining is a new approach to deal with the fast rate of growing of information. With this rapid growth of high dimensional data in many domains (e.g. machine learning, pattern recognition, customer relation management and data mining), attribute reduction is considered as a necessary task of preprocessing techniques for learning process to extract rule-like knowledge from information system [1]. In recent years, the goal of many researches is to combine data mining algorithms with information visualization methods to use the advantages of both approaches [2-6] . Interaction is important for useful visual data mining. The data analyst must be able to interact with the presented data and mining parameters according to the requirements. Attribute reduction is a process of eliminating the