Exploring Social Media Trends - A Kannada Dataset Analysis

Platforms for social media are dynamic tools that help ideas spread and develop quickly. This study introduces a novel way of locating popular Kannada-language subjects on social networking sites. This study also involves data preprocessing, feature extraction, and data visualization approaches to reveal underlying patterns and insights using a sizable dataset made up of Kannada text, tweets, hashtags, and news headlines. This method efficiently incorporates both sophisticated Machine Learning models, such as N-grams and word tokenization, and Deep Learning models, including sentence transformers and U-map embeddings. The examination of coherence and silhouette scores is used to validate the models. The main goal of this research is to offer an in-depth analysis of issues that regularly come up in Kannada debates on social media. This enables organizations, researchers, and content producers to make well-informed decisions, comprehend user opinion more thoroughly, and keep up with rapidly changing technological developments. In essence, this study helps provide a thorough understanding of the constantly changing digital ecosystem. © 2023 IEEE.