OPINION MINING FUNGSI KPI (KEY PERFORMANCE INDIKATOR) DENGAN ALGORITMA NAÏVE BAYES CLASIFIER DAN SUPPORT VECTOR MACHINE (SVM)
Abstract
Performance is measured to determine the extent to which goals are realized so that management can act quickly in making decisions. The performance measure for each agency is a key performance indicator (KPI). The research was conducted using the CRISP-DM (Cross Industry Standard Process Model for Data Mining) methodology. Data is collected on Twitter by taking Opinion Tweets using KPI (Key Performance Indicator) and will be analyzed using RapidMiner with the Naive Bayes algorithm and Support Vector Machine (SVM). The research results obtained accuracy using the SVM algorithm with SMOTE showing a result of 72.32% higher than the Naïve Bayes algorithm which obtained a result of 60.95%. The AUC of the Naïve Bayes algorithm with SMOTE is 0.875 and the AUC of the Support Vector Machine (SVM) algorithm with SMOTE is 0.772. From these results it can be seen that the SVM algorithm will predict better than the Naïve Bayes algorithm in this study to analyze sentiment in the use of KPI (Key Performance Indicator) functions.