How Social Media Shapes Public Opinion: A Closer Look at the 2024 Platform X Case

In today's world, social media plays a huge role in shaping what people think. It's not just about sharing photos or chatting with friends anymore. Social media has become a powerful tool that influences public opinion in big ways. This is why predicting how public opinion changes is so important for governments and companies. They need to understand these changes to make better decisions. One way to predict these changes is by using something called the BRT model. This model looks at how people connect on social media and how important certain topics are over time. It helps to identify what people are talking about with great accuracy. But how does it work? Well, it treats topics like players in a game, where each topic competes with others. This is where the TEP-BS model comes in. It uses something called Stochastic Competitive Learning to predict how these topics will evolve. Now, you might be wondering how well this model works. In September 2024, it was tested on four trending topics on Platform X. The results were impressive. The TEP-BS model performed significantly better than other methods. It had an average increase in F1 score of 60%, 42%, 36%, and 44% compared to TF-IDF-LDA, NMF, HDP, and GAT respectively. This means it was much more accurate in predicting how public opinion would change. But why is this important? Well, understanding public opinion can help governments and companies respond better to people's needs and concerns. It can also help them make decisions that are more in line with what the public wants. However, it's also important to think critically about these models. They are not perfect and can sometimes make mistakes. It's always good to question the data and the methods used to analyze it. In the end, social media is a powerful tool that shapes public opinion in big ways. Models like TEP-BS can help predict these changes, but they should be used with caution. It's important to understand their limitations and always question the data.