Friday 21 March 2025
A team of researchers has developed a new model that can predict how much engagement a social media post will receive, just hours after it’s been published. The model, called IC-Mamba, uses a combination of historical data and machine learning algorithms to forecast how many likes, shares, and comments a post will get.
The researchers tested the model on four different datasets, each containing thousands of social media posts related to different topics, such as natural disasters, climate change, vaccines, and COVID-19. They found that IC-Mamba was able to accurately predict engagement levels for all four datasets, with an average error rate of just 12%.
One of the key features of IC-Mamba is its ability to handle interval-censored data, which means that it can account for the fact that social media users often engage with posts at irregular intervals. This is particularly important in the early stages of a post’s life cycle, when engagement patterns are still forming.
The researchers also found that IC-Mamba was able to accurately predict engagement levels for different types of posts, including those related to misinformation and disinformation. This could be an important tool for social media platforms looking to identify and remove harmful content from their sites.
IC-Mamba is not a simple model – it uses a combination of complex algorithms and statistical techniques to make its predictions. But the researchers believe that it has the potential to be a powerful tool for understanding how people engage with social media, and for helping to promote healthy online behaviors.
The team’s findings were published in a recent paper, and they are already being used by researchers and policymakers around the world. The model could potentially be used to help identify and address issues such as online harassment, propaganda, and misinformation, which can have serious consequences for individuals and society as a whole.
Overall, IC-Mamba represents an important step forward in our understanding of how people engage with social media, and it has the potential to be a powerful tool for promoting healthy online behaviors.
Cite this article: “Predicting Social Media Engagement: A New Model Breakthrough”, The Science Archive, 2025.
Social Media, Prediction, Engagement, Machine Learning, Algorithms, Data, Model, Ic-Mamba, Online Behaviors, Misinformation
