PENERAPAN MODEL SUPPORT VECTOR MACHINE TEXT MINING PADA KOMENTAR REVIEW SMARTPHONE ANDROID VS BLACKBERRY DENGAN TEKNIK OPTIMASI GENETIC ALGORITHM

Abstract

The smartphone market is now more and more, not only in the conventional penjualanya but have penetrated in the online shop. But not all smartphones have good quality to support the needs of consumers and it is to be noticed by the consumer. Before consumers decide to buy android smartphone or belackberry consumer should know the details of the specifications and functions of the smartphone, it can be learned from the testimony and opinion or the results of a user review comments or belackberry difference android smartphone. Reading the comments of the review as a whole can be time consuming, but if only a few comments are read reviews on evaluations will be biased. Of some of the techniques most often used for data classification is Support Vector Machines (SVM). SVM has the advantage of being able to identify the separate hyperplane that maximizes the margin between two different classes. Selection of features at once setup parameters in SVM significantly influence the results of classification accuracy. Therefore, in this study used the incorporation of feature selection methods, namely Genetic Encryption (GA) in order to improve the accuracy of the classifier Support Vector Machines. This research resulted in the classification of text in a positive or negative form of product reviews Smartphone. Measurement is based on Support Vector Machines accuracy before and after the addition of feature selection methods. While the measurement accuracy is measured by the confusion matrix and ROC curves. The results showed an increase in Support Vector Machines accuracy of 71.00% with the addition of Genetic Encryption be 78.02%

Keywords : Comments , Review, SVM , Genetic Encryption ( GA ) , Confusion Matrix , ROC curve