<![CDATA[This research identifies a research gap in understanding the impact of contextual factors on sentiment and toxicity within online discussions of sports events, focusing on the MotoGP event in Mandalika. By exploring how contextual nuances influence public sentiment and toxicity levels, this study aims to enhance the effectiveness of online discourse management and improve user experiences in digital platforms hosting event-related content. This research investigates the nuances of public sentiment and toxic language within textual data, specifically focusing on content videos of the MotoGP event held in Mandalika. Methodologically, the study embraces the CRISP-DM framework, facilitating structured data analysis, model development, and subsequent deployment. The findings reveal promising outcomes in terms of the performance of machine learning algorithms; notably, the k-NN algorithm attains an accuracy rate of 94.33%, precision of 96.48%, recall of 92.01%, f-measure of 94.19%, and an AUC score of 0.982. Similarly, the Support Vector Machine (SVM) demonstrates commendable accuracy, achieving 87.54%, precision of 99.53%, recall of 75.45%, f-measure of 85.82%, and an AUC score of 0.986. Furthermore, the toxicity analysis uncovers varying levels of harmful language, ranging from 0.01229 to 0.08933. These findings underscore the imperative nature of considering both sentiment dynamics and toxicity in managing online discourse effectively and enhancing user experiences across digital platforms. The sentiment analysis underscores the importance of understanding and effectively managing public emotions in the context of sports events like MotoGP. By acknowledging and addressing positive and negative sentiments, event organizers can better engage with their audience, mitigate potential issues, and ultimately enhance the overall experience for all involved.]]>
Copyrights © 2024
