Your browser doesn't support the features required by impress.js, so you are presented with a simplified version of this presentation.

For the best experience please use the latest Chrome, Safari or Firefox browser.

Use a spacebar or arrow keys to navigate

Close ties
and large networks

A small world, really?

Christophe Prieur,
Télécom ParisTech, dept. SES, I3

École thématique Réseaux

Oléron 2015

22 sept 2015

4 degrees of separation

Backstrom et al., 2012

Followup on Milgram, 1967, The small world problem

Small world, close horizon

  • Random graphs do have very small distances
  • A path of length 5 is looooong
  • (btw, Facebook's diameter is ~40)

two regimes of
phone communication

Licoppe, Smoreda, 2000, 2005
connected presence many short calls, sms
bridging distance long calls

distance matters...

probability of a call
avg. duration

Lambiotte et al., 2008

... on network structure

contactposition in patterns
with highest numer of calls central
with longest calls intermediate
random peripheral

Stoica & Prieur, 2009

Commenting on Flickr

  • Web platform to share photos
    (beach, sunset, cats, dogs, street, etc.)
  • Many ways to browse through photos
    (tags, groups, users, "interestingness" ranking, etc.)
  • Ties between users
    contact lists, comments on each other's photos, etc.

Length of new ties

short (2)
long (3+)


repeatedshort (2)long (3+)
76 %17 %7 %
  • all (40M) comments
  • on all (10M) public photos
  • (between 3M users)

Raux & Prieur, 2011

The filter bubble

Pariser, 2011
Browsing by affinity
+Recommendation algorithms
=Self-reinforcing process

Encouraging diversity

Graells-Garrido, 2015

  • UX studies confirming the automaton bias
  • experimental designs to show diversity

    clustered items, treemaps...

Diversity makes good matches

Backstrom & Kleinberg, 2014
about couples in Facebook ego-networks

  • predicts if ego is
    "in a relationship"
  • if yes, guesses the lover with 50%

    (only 25% when using shared friends)

  • when bad dispersion value, twice as much likely to break within 2 months


  • Given ego and alter,
  • take ego's network without ego and alter,
  • dispersion (u,v)
    = sum of distances between all pairs of nodes
  • distance (u,v)
    = 1 if u and v (friends of) friends of each other,
    0 otherwise

Escaping your network:

Online dating

Bergström, 2014

  • 10% of respondents have used a dating platform,
    1/3 among 18-25

    (Ined, 2006)

  • 24 days (median) between 1st interaction and sex

    (while >1 month for more than 3/4 couples outside dating platforms)

« Ça peut être compliqué aussi de draguer dans des soirées parce que c'est des amies et des amies d'ami(e)s. Y'a toujours des histoires et ça pose pas mal de problèmes [...] En fait ça complique vachement les choses. Sur internet, c'est plus simple de rencontrer des personnes qui sont vraiment nouvelles, qui sont extérieures à tout ça. »
(Paul, 26 ans)

Tinder: far, but close

  • based on Facebook account
  • photo(s), first name, age
  • geographical proximity (shown)
  • common interests (likes, shown)
  • shared friends of friends (not shown)
  • shared friends (shown)

    → double-sided information