A Supervised Local Linear Embedding Based SVM Text Classification Algorithm

In order to solve the problem of high dimension in text classification, this paper imported local linear embedding algorithm for dimension reduction. However, the original LLE did not necessarily make the loss of information minimize in process of reduction, so we combinated its two loss function together and improved it firstly. Then, linked the improved LLE and supervised learning and support vector machine algorithm together, so this paper proposed a supervised local linear embedding based SVM text classification algorithm. Finally, we designed three experiments for comparing, and the results of experiments indicated the algorithm could be used for dimension reduction effectively, and it did really improve the accurate rate in text classification.