Classification of Text Mining Review Oil Disfusser Products Using Naive Bayes Classification: Classification of Text Mining Review Oil Disfusser Products Using Naive Bayes Classification

Abstract

Health is the main thing, especially when an outbreak of virus spreads and worries and the possibility of stress increases. Everyone wants to live healthy and avoid stress, that's why the use of natural medicines is the choice of one of them using essential oils or known as aromatherapy. The use of essential oils that are turned into steam and inhaled can produce a calming effect, the function of the pulse is more regular so that it is relaxed and fresher. Is a diffuser which is a tool to convert oil into steam, the use of the diffuser is rife and sales also get negative and positive comments from consumers. This has led researchers to examine consumer opinions about the diffuser product. Using the Naïve Bayes Classifier method to classify reviews based on positive sentiment class and negative sentiment class. From the labeling results, it is seen the association of texts in each sentiment class to find information that is considered important and can be useful for decision making. The classification results using the Naïve Bayes Classifier model obtain an accuracy rate of 76.00% and the value of the split ratio reaches the level of accuracy. The results of this study are expected to be developed and contribute to the development of sentiment analysis research by applying different methods.