Asian Journal of Pure and Applied Mathematics

Sentiment analysis, a fundamental task in natural language processing, plays a pivotal role in deciphering the emotional tone conveyed within textual content. In this study, we delve into the realm of sentiment analysis, focusing specifically on the application of a Gated Recurrent Unit (GRU) model  to the 2016 Donald Trump tweets dataset. By employing the GRU model with carefully chosen parameter settings, we aim to unravel the sentiment dynamics inherent in the dataset. Subsequently, we present the outcomes of our experiment, where we assess the model's performance using critical  metrics such as accuracy, precision, and recall. Notably, our findings reveal an accuracy rate of   92%, attesting to the model's proficiency in accurately categorizing sentiments. Moreover, precisionand recall both achieve impressive scores of 94%, underscoring the model's adeptness in recognizing  both positive and negative sentiments. Through this investigation, we illuminate the capability of the  GRU model to conduct sentiment analysis effectively, shedding light on the emotional undercurrents within the 2016 Donald Trump dataset and enhancing our understanding of sentiments during the election period.