Summarizing Netizens' Sentiments Towards the 1st Indonesian Presidential Debate using Lexicon Sentiment Analysis

9th Annual Basic Science International Conference 2019

自然科学（总论）

Lestari, Ariesta^1 ; Karolita, Devi^1

Departement of Informatics Engineering, Faculty of Engineering, Palangka Raya University, Indonesia^1

关键词: Indonesia;    Logistic regressions;    Media communications;    Prediction model;    Predictive modeling;    Presidential election;    Public engagement;    Social media platforms;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/546/5/052041/pdf
DOI  :  10.1088/1757-899X/546/5/052041

学科分类：自然科学（综合）

来源: IOP

【 摘 要 】

Twitter is one of the popular social media platforms in Indonesia. This platform has been used as a media communication and public engagement tool for many purposes, especially in political and governance domains. During the process of 2019 Indonesian Presidential Election, many people use Twitter to express their opinion/sentiment towards the election process. In this paper, we investigate the nature of people's opinion towards the Indonesian Presidential Election after the 1st debate. The goal of this study is to perform exploratory sentiment based analysis of Twitter data, and that was gathered after the 1st debate. We used lexicon sentiment analysis to calculate the sentiment of political tweets collected after the 1st debate. The identification of positive and negative opinion was automatically conducted using the available dictionary. Our result shows that sentiment of the netizen towards the 1st Presidential debate was mostly negative. In addition to this result, a predictive model was generated using CART and logistic regression to predict the netizens' sentiment. This experiment shows that the accuracy of the prediction model reaches 90%. Therefore, our study suggests that Twitter data can be used to analyse citizens' sentiment toward the Indonesian Presidential Debate and can generate a model to predict citizens' future sentiment toward the next debate.