Analisis Sentimen Opini Publik Terhadap Konser Coldplay Di Jakarta Pada Twitter Menggunakan Metode Support Vector Machine

Abstract

Advances in information and communication technology are changing the way humans obtain information. One of them is through social media Twitter. Twitter is a social media that is famous as a medium for general public opinion. There are many discussion topics that people respond to on Twitter. One of the topics on Twitter social media that is trending is related to the Coldplay Jakarta concert. The Coldplay Jakarta concert event received various responses from netizens. This is because the public is very enthusiastic about watching this British music group. However, public enthusiasm did not run smoothly due to the emergence of various phenomena and problems that occurred before and after this event was held. Therefore, the sentiments given are very diverse. Starting from negative sentiments to positive sentiments. The sentiment analysis process can find and solve problems based on public sentiment on Twitter. The dataset was obtained through a data crawling process using Google Colabs with the Python programming language. The total dataset obtained was 1,831 raw datasets. After that, the dataset is processed in data cleansing which aims to remove components that are not needed in sentiment analysis. Next, the dataset is labeled negative and positive. Then pre-processing is carried out on the data that has been previously labeled, and a word weighting process is also carried out using Term Frequency - Inverse Document Frequency (TF-IDF). After that, modeling was carried out using the Support Vector Machine classification method and the final process was test evaluation. The number of datasets obtained after going through the cleaning and labeling process is 1,000. The dataset is divided into training data and test data in a ratio of 80:20. The results obtained from the Support Vector Machine method show an accuracy percentage of 80.97%, precision of 81.59% and recall of 79.12. The sentiment test results from the test data were 107 positive and 93 negative