Computerised Sentiment Analysis on Social Networks. Two Case Studies: FIFA World Cup 2018 and Cristiano Ronaldo Joining Juventus

摘要

Sentiment analysis on social networks plays a prominent role in many applications. The key techniques here are how to identify opinions, to classify sentiment polarity, and to infer emotions. In this study, we proposed a sentence-level sentiment analysis based on a microblogging platform like Twitter. Comprehensive evaluation results on realworld scenarios such as the FIFA World Cup 2018, and the Cristiano Ronaldo's transfer to Juventus in the summer of 2018 demonstrate a correlation between the polarity (negative or positive) and fan's sentiments. In addition, the evaluation of several machine learning techniques; applied to identify the polarity and related emotions, revealed that the SVM outperforms other models such as Naive Bayes, ANN, kNN, and Logistic Regression. Additional studies should be addressed to evaluate the proposed system on different sport events, and scenarios.