<![CDATA[With increasingly advanced technology nowadays, it become easier for people to shop. Previously we had to shop in person, now thanks to technology we can shop online with E-Commerce. E-Commerce provides an opportunity for producers to promote their products to be more accessible and makes it easier for consumers to share activities such as shopping and providing reviews. Reviews play an important role in a product, especially in building consumer confidence in choosing the desired product. Therefore, sentiment analysis is needed to help choose an E-Commerce application. Here the author will carry out Sentiment Analysis using Machine Learning. The purpose of this research is to conclude and identify whether a review is a positive review or a negative review. The algorithm used in this research is Naïve Bayes and Decision Tree using TF-IDF. The research results obtained are that the Naïve Bayes algorithm has higher accuracy with an accuracy of 0.88 or around 88%, and the negative label has a precision prediction result of 0.87, recall has 1.00, fi-score has 0.93, and finally with support 26, while the positive label with precision prediction results has 1.00, recall has 0.33, fi-score has 0.50, and the last one has support 6.]]>
Copyrights © 2024
