Sentiment Analysis Of Practice Service's Questionnaire Using Naïve Bayes Classifier Method
Selly Yunita, Yusni Amaliah, Suprianto Suprianto, Aida Indriani, Muhammad Noer Fadlan, Muhammad · 2021 3rd International Conference on Cybernetics and Intelligent System (ICORIS) · 2021
Questionnaire is one instrument to measure the quality of service given by a college, particularly practicum service, including laboratory assistants. The large number of students is time and energy-consuming to analize the sentiment from the questionnaire. Therefore, a sentiment classifier is required to classify the students' comments from the questionnaire. The system was built using Naive Bayes Classifier to determine the statements in questionnaire expresss a positive, negative, or neutral sentiment. The method will calculate the Vmap value (probability theory) and the statement will be classified into the class with the highest Vmap value. The calculation was started by processing the questionnaire through case folding, tokenization, normalization, stemming, and filtering, then, finally, the classification with Naive Bayes Classifier. The research shows that Naive Bayes Classifier can analyze the sentiment in a questionnaire with the highest accuracy is 83,33% (240 training data and 60 testing data), compared to 81% accuracy with 200 training data and 100 testing data.