Department Systems Analysis, Integrated Assessment and Modelling

Exposure

Photo by Francesco Ungaro from Pexels

Community detection consists of extracting the affinity between agents of a system, which
is extracted from quantities such as the frequency of interactions. When analyzing datasets,
however, the absence of a connection between agents might not be due to a lack of affinity,
but rather to the fact that these agents never met. For example, Sandra and Paul would like
each other a lot, if they only met.
We introduce exposure into community detection, as an additional mechanism to explain the
lack of links among agents. For problems in which being exposed to other agents is crucial
towards the development of an affinity, this leads to enhanced community detection. In our
example, our community detection scheme is aware that if Sandra and Paul never met, it
might be either because they are incompatible, or just because they have not met yet.

Publications

Extbase Variable Dump
array(2 items)
   publications => '30321' (5 chars)
   libraryUrl => '' (0 chars)
Extbase Variable Dump
array(1 item)
   0 => Snowflake\Publications\Domain\Model\Publicationprototypepersistent entity (uid=30321, pid=124)
      originalId => protected30321 (integer)
      authors => protected'Othman, S.; Schulz, J.; Baity-Jesi, M.; De Bacco, C.' (72 chars)
      title => protected'Modeling node exposure for community detection in networks' (58 chars)
      journal => protected'In: Cherifi, H.; Mantegna, R. N.; Rocha, L. M.; Che
         rifi, C.; Micciche, S. (Eds.), Complex networks and their applicat
         ions XI. Proceedings of the eleventh international conference on complex net
         works and their application' (255 chars)
      year => protected2023 (integer)
      volume => protected0 (integer)
      issue => protected'' (0 chars)
      startpage => protected'233' (3 chars)
      otherpage => protected'244' (3 chars)
      categories => protected'networks; community detection; latent variable models' (53 chars)
      description => protected'In community detection, datasets often suffer a sampling bias for which node
         s which would normally have a high affinity appear to have zero affinity. Th
         is happens for example when two affine users of a social network were not ex
         posed to one another. Community detection on this kind of data suffers then 
         from considering affine nodes as not affine. To solve this problem, we expli
         citly model the (non-) exposure mechanism in a Bayesian community detection 
         framework, by introducing a set of additional hidden variables. Compared to 
         approaches which do not model exposure, our method is able to better reconst
         ruct the input graph, while maintaining a similar performance in recovering 
         communities. Importantly, it allows to estimate the probability that two nod
         es have been exposed, a possibility not available with standard models.' (831 chars)
      serialnumber => protected'' (0 chars)
      doi => protected'10.1007/978-3-031-21131-7_18' (28 chars)
      uid => protected30321 (integer)
      _localizedUid => protected30321 (integer)modified
      _languageUid => protectedNULL
      _versionedUid => protected30321 (integer)modified
      pid => protected124 (integer)

Modeling node exposure for community detection in networks

In community detection, datasets often suffer a sampling bias for which nodes which would normally have a high affinity appear to have zero affinity. This happens for example when two affine users of a social network were not exposed to one another. Community detection on this kind of data suffers then from considering affine nodes as not affine. To solve this problem, we explicitly model the (non-) exposure mechanism in a Bayesian community detection framework, by introducing a set of additional hidden variables. Compared to approaches which do not model exposure, our method is able to better reconstruct the input graph, while maintaining a similar performance in recovering communities. Importantly, it allows to estimate the probability that two nodes have been exposed, a possibility not available with standard models.
Othman, S.; Schulz, J.; Baity-Jesi, M.; De Bacco, C. (2023) Modeling node exposure for community detection in networks, In: Cherifi, H.; Mantegna, R. N.; Rocha, L. M.; Cherifi, C.; Micciche, S. (Eds.), Complex networks and their applications XI. Proceedings of the eleventh international conference on complex networks and their application, 233-244, doi:10.1007/978-3-031-21131-7_18, Institutional Repository

Contact

Dr. Marco Baity Jesi Group Leader (he/him) Tel. +41 58 765 5793 Send Mail