Deep learning for Twitter sentiment analysis about the pros and cons of Covid-19 vaccines in Indonesia

The pros and cons of implementing the coronavirus disease vaccination in Indonesia could be taken into account because they sprung up remarkably and virally in social media like Twitter. Both of the pros and cons expressed on social media deserved to be analyzed by the researchers as a consideration of making government policy and to ensure that the policy was feasible to use. The technique that could be applied and combined to sentiment analysis was a word matching approach based on a lexicon dictionary (Lexicon-Based Approach). The stages of research carried out with Data Collection, Pre-processing, The Used Model, Experiment, and Model Assessment, as well as evaluation and validation of the results. The researchers crawled 1000 Indonesian language tweets data on the pros and cons of implementing Covid-19 vaccination using the R programming language with the keyword #vaksincorona and carried out a Text Cleaning Process. Then, it brought about the distribution of 472 training data and 528 testing data. Pre-processing was facilitated by Case Folding, Stemming, Tokenizing, and Stopword Removal. Using the Deep Learning Model, this study produced text classification in the positive and negative sentiments of pro and contra opinions. Based on the accuracy of the Deep Learning Model, the Confusion Matrix, and ROC Curve, the results showed that the accuracy value of the Deep Learning model was 88,83% and the AUC value was 0,955.