Near Real-Time Sentiment Analysis In Cross Domain Applications

Sports events' media consumption patterns have started transitioning to a multi-screen paradigm, where, through multitasking, viewers are able to search for additional information about the event they are watching, as well as contribute with their perspective of the event to other viewers. As such, this paper focuses on the sentiment classification of sports events related tweets that were published during the transmission of the respective events, thus enabling the understanding of the sentiment of the viewers throughout the event and, consequently, allowing sports teams and those in charge of the audiovisual production with insights on the final consumers perspective of sports events. In addition, complementing existing sentiment analysis approaches, we demonstrate that through the use of Wikipedia knowledge graph embeddings, it is possible to improve sentiment classification performances, achieving an accuracy of 94.7\%, while being able to pre-process and classify over 500 tweets per second.