ANALISIS SENTIMEN PENERAPAN GENOSE DI INDONESIA MENGGUNAKAN ALGORITMA NAÏVE BAYES CLASSIFIER DAN K-NEAREST NEIGHBOR

Hanif Manggala Putra, NIM. 17106050003 (2021) ANALISIS SENTIMEN PENERAPAN GENOSE DI INDONESIA MENGGUNAKAN ALGORITMA NAÏVE BAYES CLASSIFIER DAN K-NEAREST NEIGHBOR. Skripsi thesis, UIN SUNAN KALIJAGA YOGYAKARTA.

Preview	Text (ANALISIS SENTIMEN PENERAPAN GENOSE DI INDONESIA MENGGUNAKAN ALGORITMA NAÏVE BAYES CLASSIFIER DAN K-NEAREST NEIGHBOR) 17106050003_BAB-I_IV-atau-V_DAFTAR-PUSTAKA.pdf - Published Version Download (2MB) | Preview
	Text (ANALISIS SENTIMEN PENERAPAN GENOSE DI INDONESIA MENGGUNAKAN ALGORITMA NAÏVE BAYES CLASSIFIER DAN K-NEAREST NEIGHBOR) 17106050003_BAB-II_sampai_SEBELUM-BAB-TERAKHIR.pdf - Published Version Restricted to Registered users only Download (4MB)

Abstract

The Covid-19 virus is a kind of virus that attacks the human respiratory system and spreads quickly. This virus originated from Wuhan, China. Hence, the Indonesian government takes a role in reducing the increasing number of patients because of the Covid-19 virus by providing several early detection tools for the Covid-19 virus, one of which is GeNose. GeNose is the latest Covid-19 virus detection tool made by UGM which has been officially authorized by the Covid-19 Handling Task Force as a requirement before traveling using trains and buses. In response to the GeNose implementation in Indonesia, the public has their own opinion which is uploaded in social media Twitter posts.. This study aims to analyze Twitter sentiment towards the GeNose application in Indonesia using the Naïve Bayes Classifier and K-Nearest Neighbor methods. The data in this study used 10812 tweets using a composition of 3868 training data with labels and 6944 test data without labels. The results in this study on the learning process of 3868 training data resulted in an accuracy of 70.15% for the Naïve Bayes Classifier algorithm and 70.02% for the K-Nearest Neighbor algorithm with a value of k = 17. The implementation results of the test data were carried out as many as 6944 tweets about the GeNose application in Indonesia by utilizing the Naive Bayes Classifier classification model which resulted in 1680 positive sentiment tweets, 1146 negative sentiment tweets, and 4118 neutral sentiment tweets while the K-Nearest Neighbor classification model with a value of k = 17 resulted in 1645 tweets of positive sentiment, 806 tweets of negative sentiment, and 4493 tweets of neutral sentiment. On the other hand, the time required to perform the classification process on 6944 test data was relatively shorter when using the Naive Bayes Classifier method, which was 0.1 seconds rather than the K-Nearest Neighbor method. Meanwhile, the K-Nearest Neighbor method took a long time, which was 6.27 seconds.

Share this knowledge with your friends :

Actions (login required)

View Item