Sentiment Analysis of Hotel Reviews Using Support Vector Machine

Abstrak

With technology nowadays, everyone can leave their review about a hotel on the internet. This creates a new issue for the hotel itself because the reviews can come in in thousands amount. This will consume a lot of time to handle these reviews manually. In this study, a sentiment analysis model will be made to overcome the issue. The data in this study is collected from Kaggle website. This data contains 20,491 reviews about a hotel. The data will then be preprocessed and given a label for each data point. Then, the model is trained using the clean data. The model will use Naïve-Bayes, Logistic Regression, and Support Vector Machine algorithm. From the result performed, it's concluded that Support Vector Machine performed more accurately with 94% rate.