KOMPARASI ALGORITMA DATA MINING UNTUK ANALISIS SENTIMEN APLIKASI PEDULILINDUNGI

Abstract

The COVID-19 pandemic has caused many changes to occur in Indonesia. In the PeduliLindungi application to the community, the government makes the application to the community in the hope of being able to provide a warning if it enters the covid-19 zone and various other information from covid-19 [1]. The main purpose of this study is to analyze the sentiments of PeduliLindungi users who are currently used during the Covid-19 pandemic, where this application has begun to be used to travel anywhere and anytime to find out whether the user has vaccinated or not and various other things. such as the spread of the virus and the location of vaccination. The dataset for this study was taken from the Play Store. The algorithm used is Support Vector Machine and Naive Bayes to classify the data set. The data collection technique is Text Mining and compares the results of the two specified algorithms. The results of this research are Support Vector Machine with TF IDF Vectorizer with 89.05% accuracy followed by Support Vector Machine with Count Vectorizer, Naive Bayes with TF IDF Vectorizer and Naive Bayes with Count Vectorizer.