Exploring global Twitter sentiments on millet confections / Parminder Singh Dhillon

To assess the market potential of millet-based confections, this study aimed to elucidate the attitudes, opinions, and ongoing discussions of social media users from the top ten economies of the world. Text analysis and machine learning approaches were applied to classify the sentiments of the tweets as positive, neutral, or negative. The Naive Bayes classifier was applied to improve the precision of sentiment analysis. The process revealed the top themes presented as the top 50 phrases by their frequency in tweet data. Furthermore, the subjectivity distributions and polarity in the results provide intricate emotional perspectives. Additionally, a bar chart was used to display the distribution of positive, neutral, and negative tweets while word popularity was visualized through word clouds and word pair clouds. The insights from this research are valuable for businesses working with millets, marketers, and legislators globally.