Analisis Klasifikasi Sentimen Pengguna Media Sosial Twitter Terhadap Penundaan Pemilu Presiden Tahun 2024

Abstract

The discourse of postponing the 2024 election and extending the term of office of the president has stimulated sentiment in some regions in Indonesia. The discourse has implications for the extension of the president's term of office which is considered unconstitutional. The issue of postponing the election is clearly not the aspiration of the people, but only the interests of political passion and lust among the rulers who want to perpetuate power so that it is considered to violate and insult the constitution which stipulates that elections are held every five years. This study aims to determine how to analyze the sentiment of postponement the 2024 election on comments from the Indonesian people on Twitter or called tweets. The number of comments that will be used in the study is 1826 consisting of 710 positive sentiments and 1116 negative sentiments. The research method used is the Cross-Industry Standard Process for Data Mining (CRISP-DM) method. With this method, the stages of research carried out are data collection, business understanding, data understanding, data preprocessing, data labeling, modeling, evaluation, and deployment. Based on the results obtained at the modeling stage, the tweet data that has been collected is then processed and analyzed for sentiment with a train/test split data model and k-fold cross validation using the Naive Bayes (NB) algorithm, Support Vector Machine (SVM), Deep Learning (DL) ), and Desicious Tree (DT) and a comparison of 80:20 for training and testing data, the highest accuracy value is obtained by using the train/test split data model using the Naive Bayes algorithm to produce an accuracy of 80,55%.

Analisis Sentimen Masyarakat Mengenai RUU Perampasan Aset Di Twitter Menggunakan Metode Naïve Bayes

“…Tweet atau cuitan sendiri adalah sebuah teks yang di tuliskan oleh pengguna untuk menyampaikan opini penulis ke dalam topik terkini, dalam Tweet inilah peneliti dapat mengambil data yang bisa di analisis apakah orang ini pro atau kontra dengan topik tersebut [7]. Data Twitter tersebut akhirnya akan dianalisis menggunakan algoritma Naïve Bayes untuk mengetahui pandangan dari publik mengenai topik yang akan dibahas peneliti [8].…”