The rapid proliferation of misinformation in the digital age poses significant challenges to democratic processes, particularly in multilingual and multicultural societies like India. This thesis investigates the thematic structure, temporal patterns, and user engagement dynamics of misinformation during two general election cycles (2019 and 2024) and major socio-political events. By leveraging computational methods such as BERTopic for topic modeling and sentiment analysis using large language models (LLMs), this research analyzes data from BoomLive, a fact-checking organization, and Twitter activity from Mohammed Zubair, co-founder of Alt News. The findings reveal distinct patterns in misinformation spread, with peaks aligning closely with electoral milestones and thematic shifts reflecting political and communal nar ratives. Health-related misinformation dominated during the COVID-19 pandemic, while communal and political falsehoods surged during elections. User engagement varied across themes, with politically charged misinformation eliciting polarized reactions and communal narratives driving higher engagement metrics. Sentiment analysis highlighted users’ tendency to engage selectively with fact-checks that aligned with their ideological beliefs, underscoring the role of motivated reasoning in shaping public discourse.