Implementasi Algoritma Decision Tree dan Naïve Bayes Untuk Klasifikasi Sentimen Terhadap Kepuasan Pelanggan Starbucks

Abstract

Indonesia is included in the category of countries with the largest population in the world, this situation is a business opportunity for entrepreneurs who enter the coffee shop industry market. Researchers utilize one of the grouping methods, namely data mining classification in order to help business entities to identify different groups in the Starbucks customer satisfaction database. The purpose of this research is to be able to group categories into 3 classes, namely satisfied, quite satisfied and dissatisfied using the Decision Tree & Naive Bayes algorithm. So that it can find out public opinion on Starbucks customer satisfaction, in this study the aim was to obtain accuracy, precision and recall values and find out the best algorithm for data mining classification of Starbucks customer satisfaction. In this study using test data obtained from tweets with the keyword "Starbucks" from Twitter. The results of this study where the sentiment classification process for Starbucks customer satisfaction obtained a neutral category, it can be seen from the reviews using the keywords "starbuck OR starbucks OR #starbucks "The results obtained were positive comments of 476 tweets with a percentage of 19.2%, neutral comments of 1743 tweets with a percentage of 70.3% and negative comments of 258 tweets with a percentage of 10.4%, so that conclusions can be drawn based on the polarity calculation, the comments on stabuck have a satisfied category.In this study, it can be concluded that the performance of the Decision Tree algorithm is better than the Naive Bayes algorithm, as can be seen from the following explanation.The Decision Tree algorithm results in an accuracy of 83%. Naïve Bayes on value accuracy results by 74%.