<![CDATA[Pospay application is a form of financial technology belonging to Pos Indonesia. Pospay application on the Google Play Store has more than 25 thousand user reviews. The more user reviews, the more difficult and longer it will take for prospective users and application managers to conclude information on user sentiment trends that are useful in making decisions about application use and evaluation. Sentiment analysis is the solution to this problem because sentiment analysis is able to classify unstructured data to generate sentiment information efficiently by applying data mining algorithms. This research uses the Knowledge Discovery in Database (KDD) method, where at the data mining stage, the Support Vector Machine algorithm is applied in making the model. Grid search was applied to test 3 scenarios so that the proportion of data distribution with the best accuracy was obtained, namely, 90:10, using RBF kernel with parameters: c = 1, ℽ = 1. The results of this research were a model with 95% accuracy, 91% precision, 100% recall, and 95% f1-score. Information was also obtained that the sentiment of Pospay application users on the Google Play Store tends to be positive (54.1%) but not much far from the percentage of negative sentiments (45.9%).]]>
Copyrights © 2023
