Data mining algorithms are used for analyzing data from different sources and extracting useful information from a large volume of data. Algorithms of data mining are used in E-commerce companies to help them identifying online customer behavior to recommend appropriate products based on customers’ needs. In this paper, our aim is enhancing the result of the classification techniques that applied to an online shopping agency dataset by using clustering techniques which applied to this dataset before entering it to classification techniques, so farthest first, expectation maximization (EM), and K-mean clustering algorithms are applied to an online shopping agency dataset to allocate related objects into the same cluster. After applying clustering algorithms, a group of data mining classification algorithms such as Bayes net, Naïve Bayes, K star, filtered classifier, decision table, J48, and JRIP are applied to the three clustering algorithms. A logistic model tree (LMT) classification algorithm is applied also to measure the performance parameters for each classifier. The experimental results achieved high rates in accuracy, precision, recall, Fmeasure, and ROC when compared to recent research paper