Back to top

Conclusion

  • Introduction
  • Binary Logistic Regression
  • Propagation Graphs
  • User Clustering
  • Conclusion

    • Conclusion

    This project investigates the behavioural mechanism driving information propagation in social networks, analysing and contrasting the pathways of broadcast versus viral cascades.

    By evaluating a multitude of Twitter data, we quantified how interest similarity, structural position of a retweeting decision, and depth affect retweeting behaviour. We identified the dominating role of social relationships during primary diffusion while discovering the vital role of interest similarity in later stages. This outcome connects theoretical contagion structure with empirical statistics, providing a comprehensive understanding of virality.

    In addition to quantitative results, the project offers technical strategies for marketing and public campaigns on social media, enabling statistically supported control of information spread by optimising user-targeted content based on depth potential and community cohesion.