Perbandingan Algoritma Naïve Bayes dengan K-Nearest Neighbor Untuk Analisis Sentimen Aplikasi InDrive di Playstore

Abstract

Transportation is already part of the main needs in moving from one place to another. One of them is land transportation which is the most widely used in daily needs. There are various technology companies that are competing to create online-based transportation, one of which is inDrive. Of the several online transportation services, inDrive has a different system that can negotiate or bargain for transportation rates directly. it is interesting to do an analysis to find out whether the system is worth maintaining or it will get constructive criticism so that in the future it can improve the quality of the inDrive application. Review or comment data taken from google play through google colab as much as 1200 data and processed using RapidMiner. Testing is carried out in two stages, namely training data and testing data, training data that is greater than testing data will affect the accuracy of a method. The purpose of this research is to see various positive and negative reviews in the inDrive application and make a comparison between the Naïve Bayes method and K-Nearest Neighbor with the results of 97.50% accuracy, 92.71% precision and 100% recall for Naïve Bayes. While accuracy 83.21%, precision 85% and recall 57.30% for KNN, from these results it can be concluded that the Naïve Bayes method has superior accuracy in making classifications.