<![CDATA[This research addresses the complexities of digital content analysis, focusing on toxicity, sentiment, and social network dynamics, employing the CRISP-DM (Cross-Industry Standard Process for Data Mining) as the overarching framework. The research problem centers on understanding the prevalence of toxicity, discerning sentiment nuances, and unraveling viewer interactions within social networks. Comprehensive toxicity analysis was conducted, revealing specific scores for toxicity attributes and a prevalence of positive sentiment (72.5%). Sentiment classification utilizing the k-NN algorithm achieved exceptional accuracy (98.06%), showcasing its efficacy in sentiment discernment. Social network dynamics were examined, uncovering key metrics such as Diameter (3), Density (0.002140), Reciprocity (0.000000), Centralization (0.393200), and Modularity (0.552200), shedding light on network structures and interactions. Findings underscore the need for nuanced content moderation strategies and highlight the importance of fostering positive interactions in digital spaces. Recommendations include implementing targeted moderation policies, leveraging sentiment analysis for audience engagement, and fostering community-building initiatives to promote healthier online environments.]]>
Copyrights © 2024
