Sistem Informasi Akademik Mahasiswa Universitas Brawijaya (SIAM UB) is an academic service belongs to University of Brawijaya that is used for college academic needs service. The topic about SIAM UB sometimes become trending topic at Twitter before the new semester starts. Twitter is a social media service that quite famous among citizen for giving opinion or thought through certain topic including SIAM UB. This research try to analyze some tweets about SIAM UB by classifying a tweet into the positive sentiment class or negative sentiment class. The classification process implemented on RapidMiner. The method used on this classification process is K-Nearest Neighbor and Chi Square method for feature selection. There are four main processes for the classification, which are preprocessing, term weighting, feature selection, and classification. The highest accuration score from the classification process is 86%. That accuration score was obtained when using K = 3 and using 100% feature. The percentage of the number of features which is used affects the accuracy value where the lower feature ratio is used, the accuration score became lower too.
Copyrights © 2020