Implementasi Metode CNN Pada Klasifikasi Sentimen Terhadap Pelaksanaan Piala Dunia U-17

Abstract

The World Cup is the world's largest tournament and is organized by the Federation International de Football Association (FIFA). Football is one famous sport in Indonesia; after the information that the U17 World Cup will be held in Indonesia, of course, many people talk about this, and it become a trending topic on Twitter. One way to analyze and process texts is through sentiment analysis. The purpose of this sentiment analysis is to identify positive, neutral, and negative sentences that originate from Twitter comments. The study also aims to evaluate the effectiveness of the Convolutional Neural Network (CNN) model in classifying sentimental sentences into positive, neutral and negative categories. The research implemented the CNN method. Tweet data was obtained through Google Collab. Sentiment analysis using CNN algorithms involves the results of two different data labelling methods. The results of the sentiment analysis with the CNN model using TextBlob labelling, showed 367 positive tweets, 253 neutral tweets and 48 negative tweets, with an accuracy of 89.67%, precision of 90.18%, recall of 89,67% and F1 score of 89.00%. The sentiment analysis results with the CNN using Vader Sentiment labelling, showing 80 positive tweets, 582 neutral tweets and 6 negative tweets, with an accuracy of 97.01%, precision of 96.21%, recall of 97.01% and F1 score of 96.46%. Results of sentiment analysis using the CNN model and 2 data labelling methods resulted in the following accuracy: the accuracy Model CNN with the textBlob data and the Vader Sentiment labelling were 89% and 97%, respectively.