From the past few years, social media has played a pivotal role in understanding public sentiment on contemporary pressing issues, offering real-time and large-scale collective insights. In this paper, we present a dynamic and interactive sentiment analysis platform, which is generic in terms of the number of tweets processed and the time frame of analysis. Specifically, we leverage Twitter as a key social media platform to gather and analyze over 3 million tweets, thereby mapping and visualizing the shifts of public sentiment across the United States. Built using R and the Shiny web framework, our platform displays emotion-specific sentiments (e.g. anger, joy, trust) on a U.S. state-level choropleth map, providing keyword-based searches and employing advanced text-processing techniques for robust analysis. The results indicate that certain topics (e.g. “inflation”, “election”) often correlate with strong sentiments like anger, fear, and trust, while non-political themes (e.g. “community”, “soccer”) generally elicit more positive emotions. By analyzing state-level sentiment distributions, our platform underscores significant regional variations, offering valuable insights into policy-making, marketing, and public discourse. The software can be utilized to understand and predict public sentiment trends, facilitating evidence-based decision-making in both governmental and commercial spheres.