Spreading of viral infections in social networks

(download Grapheur sample file)



The data is a small subset of that considered in a study showing how the information dynamics shape the sexual networks of Internet-mediated prostitution. Each row contains the data of a person in a social network, with the following columns:
  • Name: The ID of each person.
  • Time: Current time in the simulation
  • Sex: Female/Male
  • Infected: Parameter describing if the person is infected or not at the given time.
Furthermore, the relationships in the network (the edges between two nodes) describe sexual contacts between people.

Objectives of data mining and visualization:

  • To visualize the spreading of a viral infection trough a network of people. In this case the infection is caused by sexual contacts between people. Very similar and less harmful viral effects are observed for diffusing reputation (about people or products), for spreading gossips, for fashionable habits, etc.
  • To study the correlation between the number of connections (contacts) and the probability of being infected in time.

Grapheur sample visualization: Similarity Map and Sweep trhough time

In the figure, a plot of the network. An edge visualizes a sexual contact between the considered persons.
The color code indicates if a person has been infected, while the shape is related to the gender.
The demo file permits to activate a Sweep of the network through time, visualizing the dynamics. Grapheur Reactive Business Intelligence: sample visualization

Download the Grapheur-ready data file: viral-infection.rbi

References: The data have been collected from the dataset provided by Luis Enrique Correa da Rocha, http://www.tp.umu.se/~rocha . The original dataset was first published on:
"Information dynamics shape the sexual networks of Internet-mediated prostitution", L. E. C. Rocha, F. Liljeros, and P. Holme; Proceedings of the National Academy of Sciences of the USA 107 (13), 5706-5711 (2010).
The use case was inspired by the following article:
"Simulated epidemics in an empirical spatiotemporal network of 50,185 sexual contacts", L. E. C. Rocha, F. Liljeros and P. Holme; PLoS Computational Biology 7(3) e1001109 (2011).
Note that the spreading of the epidemic was simulated with a simpler model than the one explained in the article, in order to maintain the use case as simple as possible.