Analisis Sentimen Dengan Klasifikasi Naïve Bayes pada Review Hotel Tripadvisor

Abstract

In this era of rapidly growing data digitization, people are facilitated by the existing reviews to decide something. Reviews about something positive, of course, will also have a positive impact on those being reviewed. In this study, the tripadvisor hotel reviews dataset is taken from the Kaggle dataset. The classification method used is nave Bayes. The first thing to do after dataset retrieval is data preprocessing such as cleaning data by changing text to lowercase, eliminating numbers, eliminating double spaces, tokenization, eliminating stopwords and lemmatization. From the results of the classification with nave Bayes, the results obtained with an accuracy of 95.6%, which means that the results of the predictions are very good. For future research, other classification methods will be used to compare and find the best accuracy results.