Abstract:
The purpose of this study is to investigate the political polarization factor using Twitter data. It comprehensively explores an extensive dataset encompassing a trove of over 180,000 recent Twitter tweets, spanning the dynamic period from late 2022 to early 2023. The overarching objective of this endeavour is to unearth profound insights into the prevailing user behaviours and engagement paradigms that define the Twitter platform’s landscape. In the pursuit of these insights, we adopt a multifaceted analytical approach, wielding both qualitative and quantitative methodologies to distil knowledge from the intricate tapestry of tweet content, metadata, and the intricate network structures that underpin the Twitter ecosystem. Our rigorous investigation commences with a meticulous statistical analysis of tweet length distributions. This analysis unveils a fascinating phenomenon where a significant cohort of users opt for succinct, minimalist tweets, often comprising a solitary word. Their intent becomes evident, as these tweets appear to be strategically tailored to ride the waves of viral trends, rather than serving as conduits for substantive discourse. In parallel, our scrutiny of user engagement metrics, encompassing likes and retweets, further substantiates this observation, with the majority of tweets amassing only modest interaction. It is only a minuscule fraction of tweets that attains the coveted status of going viral. The exploration of tweet content constitutes a pivotal facet of our analytical voyage. To elucidate the semantic tapestry of tweets, we harness the power of topic modeling techniques. These techniques orchestrate the clustering of tweets into coherent themes and conversations. The resulting tapestry reveals a vibrant spectrum of major themes, with news and politics occupying a prominent position, often intertwined with links to articles and discussions concerning current events. Simultaneously, an effervescent tapestry of fan communities emerges, passionately rallying around the realms of pop culture, celebrities, and artists. These themes collectively paint a vivid portrait of the diverse discourse that Twitter encapsulates. Adding another layer of nuance, sentiment analysis accentuates our findings, with the predominant emotional tone pervading tweets being one of neutrality or positivity. However, it is worth noting that pivotal sociopolitical events tend to precipitate spikes in negative sentiment, underlining the platform’s sensitivity to the external socio-cultural landscape. Intriguing insights continue to surface as we delve into network analysis. This dimension provides a window into the intricate web of interconnected communities that coalesce within the Twitterverse. The cartography of user interactions unveils tightly knit clusters, coalescing around shared interests, common hashtags, and responses to linked content. Yet, amidst this vibrant tapestry of connectivity, a discernible fragmentation persists, with a substantial portion of users operating within confined, insular sub-communities. In culmination, this paper amalgamates an array of analytical approaches, encompassing content analysis, sentiment analysis, statistical modeling, and network analysis, to construct a panoramic understanding of user dynamics within the intricate tapestry of Twitter. These findings illuminate the prevailing motivations and collective dynamics that underpin the fast-paced, cacophonous milieu of Twitter conversations. Our work not only serves as a compass for navigating the rich, multi-faceted terrain of social data but also charts a course toward an enriched comprehension of the ever-evolving world of digital discourse.
