<![CDATA[SAMBAT Online is an application system used to accommodate complaints from the public against to the government of Malang. The incomplete features of SKPD selection related to the system made it difficult for Diskominfo of Malang City to report the complaint to the related SKPD. This is because the complaint grouping based on related SKPD is still done manually. Therefore, a system that can group complaints based on the relevant SKPD is required for time efficiency. NW-KNN is classification method which can be used to handle balanced issues that work by involving all training data in the process. The feature selection techniques that will be used are information gain and genetic algorithm to get a small number of features and high f-measure. Stages performed in the system get the best features of the first is pre-processing data, second is feature selection by using information gain, and the third is selection features by using genetic algorithm. The results of the tests performed resulted 0.22 in average of f-measure for unbalanced data and 0.39 for balanced data. These results have increased up to 0.04 for unbalanced data and 0.22 for balanced data from classification results without using feature selection process. Based on these results, it can be concluded that the classification using NW-KNN and information gain-genetic algorithm feature selection can be used to improve the classification results.]]>
Copyrights © 2019
