Euro CSS

The symposium’s program is here.

Keynote 1| Kristina Lehman

  • The friendship paradox is the phenomenon for which”most people have fewer friends than their friends have, on average”. The paradox comes from the long-tail distribution of number of nodes in a network. The network structure explains paradoxes. but what kind of structure? researchers answered that question by looking at: 1k structure which captures degree distribution, 2k structure which captures degree assortativity, and 3k structure which captures the correlation of neighbours’ degree.
  • The majority illusion: “most of my friends have more exciting lives than I do”. This illusion is a perception and can be engineered! For example, it has consequences on targeting. In this paper, researchers inferred “the economic well-being of individuals through a measure of their location and influence in the social network”. They add: “we carry out a marketing campaign that shows a threefold increase in response rate by targeting individuals identified by our social network metrics as compared to random targeting.”
  • Simpson’s Paradox suggests that “a trend exists in aggregate data but disappears or reverses when data is disaggregated by groups”.
  • As a consequence of Simpson’s paradox, reputation rates are more robust than reputation
  • Conclusion

SLIDES here


Keynote 2| Keith Payne

  • Everybody thinks to be above average (mathematically impossible). That is because we all find painful to be below average.
  • As our needs raise, so do the risk we are willing to take. In high inequality societies, people will feel they need more and will take more risks.

(I wonder whether inequality is associated with economic growth)

Google search index corpus was used to build a country index for “risky googling” (in the three categories of financial, sexual, drug & alcohol). And it turns out that income inequality is associated with risky googling.