Dynamical models of the interaction between spreaders of and exposed to truth and rumor

• Main Authors: Feria, Miguel, Oliva, Michael
• Format: text
• Language: English
• Published: Animo Repository 2018
• Subjects: Physical Sciences and Mathematics
• Online Access: https://animorepository.dlsu.edu.ph/etd_bachelors/18587
• Institution: De La Salle University

Online social media allows individuals and organizations the power to spread their influence. It is important to study how information spreads, given the effects and impact it brings to the society. Past studies investigated the spread of rumors without considering the influence of the spread of truth. We, however in this study, consider both spreader for truth and for rumor and investigate how their presence aspects the spreading ability of one another. Based on this approach, we propose two models that study the dynamic interactions of spreader-spreader and exposed-spreader. These models are built based on the epidemiological SEIR model. Then, we analyzed these models on a macro-level, where we investigated the general effects of changing the rates of spread on the whole population, and on the micro-level, where we inspected the specific effects of changing the initial spreaders. For our analysis on the macro-level, we found the reproduction numbers for truth and rumor, determined the stability at the equilibrium points, and learned the best approach for spreading truth and halting rumors. As expected, our results show that having greater rates for the spread of truth compared to rumors prevents rumors from spreading. However, we have discovered that it is enough for these rates of the spread of truth to only be slightly greater than the rates of the spread of rumors as any further increase will result in diminishing returns. In the micro-level, we considered the network topology of the population by implementing a discrete stochastic version of the model. By analyzing the node degree and betweenness centrality of the nodes (taken as individ- uals) in a network, we are able to model the importance of an individual spreader of truth or rumor in a network, and show that spreader nodes with greater degree and betweenness centrality are more effective at spreading information.