Publication Details
Abstract
The fast evolution of digital platforms has changed the way public institutions communicate with the citizens and the real-time feedback and analysis of the public opinion has become more and more important. This research questions how AI-based sentiment analysis can enhance the effectiveness of the social media communication campaign among the population in the United States through the analysis of the massive discourse on X (formerly Twitter). Based on the 2024 U.S. Presidential Race on X: Sentiment and Trends dataset, this study will discuss the opportunities and threats of using artificial intelligence and natural language processing (NLP) technology to monitor, identify and analyze the public sentiment in regard to political and policy-based communication campaigns in a highly dynamic election year. The dataset is sentiment-labeled tweets, timestamps, candidate and party affiliation, and the engagement dimension of likes and retweets, which allows conducting a multidimensional analysis of public feedback. Following the processing of data, such as text normalization, removal of noise and tokenization, supervised machine learning models are then used to label the sentiment as positive, negative, and neutral. The analysis also examines the trend of sentiment over time, sentiment differences between candidates and the dependence between sentiment polarity and user engagement. Results show that AI sentiment analysis can offer practical feedback on how messages on the communication front are perceived and enhanced by the masses on social media. Positive sentiment is typically linked to increased levels of engagement whereas increase or decrease of sentiment with time generally aligns with significant campaign announcements and policy debates. These findings help to point to the opportunities of AI-based analytics to assist in data-driven decision-making, message optimization, and adaptive communication strategies in governmental circles. This study is relevant to the literature on AI usability in open communication because it proves how sentiment analysis can be used to increase the responsiveness, openness, and confidence between institutions and citizens. The suggested framework can be generalized beyond the electoral setting to the methods of strengthening digital public communication in the contemporary context through the prism of public health, emergency communication, and policy awareness campaigns, providing a flexible solution to the mass communication issues in the modern worl8d.