Abstract:
Internet users in Indonesia reached 210 million in December 2022. Internet users can
shop through the Shopee e-commerce platform that offers ShopeePay payment features.
It is important for Shopee to know user reviews about the feature as an evaluation so
that quality can be improved. The review can be viewed on the Google Play Store. The
purpose of the study was to classify positive and negative reviews using Support Vector
Machine (SVM) with various kernel functions, including linear, polynomial, sigmoid,
and Radial Basis Function (RBF) kernels. Sentiment results are presented with a word
cloud and then viewed the text associations of each sentiment class. The polynomial
kernel provides the best classification for ShopeePay reviews with accuracy, sensitivity,
and specificity of 88.64%, 57.47%, and 98.70%, respectively. Text association of
positive sentiment of ShopeePay reviews related to vouchers, freebies, shipping, while
negative sentiment related to admin fees.