Sentiment Analysis for the Identification of Negative Situations in Soccer Matches Using Social Networks and Artificial Intelligence Techniques

This article presents an approach applying artificial intelligence techniques for sentiment analysis to identify potentially negative situations in soccer games. Two artificial intelligence techniques were employed for sentiment analysis: (1) bag-of-words and (2) computer vision. The first is used for Natural Language Processing (NLP) and sentiment identification, and the second is used for computer-based emotion recognition. Four soccer matches were analyzed using data from X social media platform (formerly Twitter). The evaluation was performed over real scenarios in Mexico: an atypical match dated March 5th, 2002; the final of the 2023 closing season; a regular match of 2024; and the game of the closing season 2024. The results indicate that for a critical event, an average negative perception of 77.5%; for a closing season match, a positive perception of 54.9%; for a regular boring season match, an average negative perception of 58.55%.; and for a final match of the closing session 2024, an average negative perception of 62.2%.