IMPLEMENTASI ALGORITMA NAIVE BAYES DALAM ANALISIS POLARISASI OPINI MASYARAKAT TERKAIT VAKSIN COVID-19

Hingga tahun 2022, Indonesia dilanda pandemi Covid-19. Covid-19 merupakan virus yang sangat mudah menyebar sehingga  Word Health Organization (WHO) menetapkan status Virus Covid-19 sebagai pandemi global. Kasus Covid-19 pertama kali terdeteksi di Indonesia pada tanggal 2 maret 2020. Sejak saat itu, banyak kasus yang terkonfirmasi positif. Pemerintah Indonesia sudah mengupayakan untuk menekan penyebaran covid-19 agar dampak negatif yang ditimbulkan covid-19 bisa dikendalikan, diantaranya dengan program vaksinasi covid-19 yang terbagi menjadi beberapa tahapan, yaitu pemberian vaksin covid-19 dosis satu, dosis dua, hingga vaksin booster. Akan tetapi, program vaksinasi covid-19 untuk masyarakat memunculkan beragam opini di jagad media sosial, terutama twitter. Opini yang disampaikan cenderung terpolarisasi menjadi sentimen yang mendukung dan menolak. Untuk itu, penelitian ini bertujuan untuk mengetahui polarisasi opini masyarakat tentang program vaksinasi Covid-19 menggunakan algoritma naive bayes. Proses dari analisis polarisasi opini meliputi pengumpulan data melalui twitter menggunakan tools rapidMiner, kemudian dilakukan preprocessing data dengan cara case folding, tokenizing, filtering dan stemming. Tahapan terakhir adalah melakukan klasifikasi opini masyaratakt algoritma naive bayes. Penelitian ini menghasilkan polarisasi opini masyarakat meliputi tweets sentimen positif sebesar 67%, tweets sentimen netral 3% dan tweets sentimen negatif 30%. Juga diperoleh tingkat akurasi dan recall masing-masing 88% dan 97%.Until 2022, Indonesia was hit by the Covid-19 pandemic. Covid-19 is a virus that spreads very easily, so the World Health Organization (WHO) has declared the Covid-19 virus status as a global pandemic. The first case of Covid-19 was detected in Indonesia on March 2, 2020. Since then, many cases have been confirmed positive. The Indonesian government has made efforts to suppress the spread of Covid-19 so that the negative impacts caused by COVID-19 can be controlled, including the Covid-19 vaccination program which is divided into several stages, namely the administration of the Covid-19 vaccine, dose one, dose two, and the booster vaccine. However, the COVID-19 vaccination program for the community has raised various opinions in the social media universe, especially Twitter. Opinions expressed tend to be polarized into sentiments of support and rejection. For this reason, this study aims to determine the polarization of public opinion about the Covid-19 vaccination program using the Naive Bayes algorithm. The process of opinion polarization analysis includes collecting data via Twitter using the RapidMiner tools, then preprocessing the data by means of case folding, tokenizing, filtering and stemming. The last step is to classify public opinion using the Naive Bayes algorithm. This study resulted in polarization of public opinion including tweets of positive sentiment by 67%, tweets of neutral sentiment 9% and tweets of negative sentiment 32%. Also obtained the accuracy and recall rates of 88% and 97%