{"title":"Classifying the Polarity of Online Media on the Indonesia Presidential Election 2019 Using Artificial Neural Network","authors":"Muhammad Afif Farisi, K. Lhaksmana","doi":"10.1109/ICoICT49345.2020.9166254","DOIUrl":null,"url":null,"abstract":"The 2019 presidential election is one of the mandatory national agendas that is covered by all of the mainstream news media in Indonesia. The function of news media as an information provider reaps criticism because they are suspected of having polarity towards certain candidates. In this paper, the polarity of news media is analyzed by performing sentiment assessment towards every news regarding each candidate. Since manual sentiment analysis is costly and time-consuming, because of the large amount of data that needs to be processed, we adopt a machine learning method to automate the sentiment analysis process. This research employs Artificial Neural Network (ANN) to classify scraped news texts from online media and TF-IDF weighting method for feature extraction. We found that the observed online media kompas.com, liputan tan6.com, republika.co.id, and tempo.co do not have significant polarity toward one of the candidates. In addition to ANN, we also compared other methods to investigate the appropriate methods for our dataset. Our experiment shows that on average, ANN obtains the best accuracy at 84.57%, compares to Decision Tree C4.5 (83.34%), Naive Bayes (SO.42%), and SVM (79.04%).","PeriodicalId":113108,"journal":{"name":"2020 8th International Conference on Information and Communication Technology (ICoICT)","volume":"42 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2020-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2020 8th International Conference on Information and Communication Technology (ICoICT)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICoICT49345.2020.9166254","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
The 2019 presidential election is one of the mandatory national agendas that is covered by all of the mainstream news media in Indonesia. The function of news media as an information provider reaps criticism because they are suspected of having polarity towards certain candidates. In this paper, the polarity of news media is analyzed by performing sentiment assessment towards every news regarding each candidate. Since manual sentiment analysis is costly and time-consuming, because of the large amount of data that needs to be processed, we adopt a machine learning method to automate the sentiment analysis process. This research employs Artificial Neural Network (ANN) to classify scraped news texts from online media and TF-IDF weighting method for feature extraction. We found that the observed online media kompas.com, liputan tan6.com, republika.co.id, and tempo.co do not have significant polarity toward one of the candidates. In addition to ANN, we also compared other methods to investigate the appropriate methods for our dataset. Our experiment shows that on average, ANN obtains the best accuracy at 84.57%, compares to Decision Tree C4.5 (83.34%), Naive Bayes (SO.42%), and SVM (79.04%).