End-of-study research internship as LIP6, Sorbonne University and CNRS (French National Scientific Research Center) lab.
- ComplexNetworks team (https://www.complexnetworks.fr/)
- Prof. Lionel Tabourier
- Prof. Fabien Tarissan
Real-world networks share non-trivial properties, such as a skewed degree-distribution, a low global density, a high local density, etc. However, existing random graph models do not capture all the properties of real world networks at the same time. In particular, the bipartite configuration model achieves in preserving a degree distribution close to the original network one, but fails in generating graphs that keeps the overlapping structures.
F. Tarissan & L. Tabourier proposed a random model [1][2] that relies on maximal bicliques to preserve overlaps in bipartite networks, by exploiting the tripartite version of the configuration model. The purpose of this study is to tackle the realistic random graph model problem, by evaluating the relevance of the tripartite model as a possible generalized answer to this issue. We will both validate and further current knowledge of this model, by examining other characteristics of real-world network.
Main classes:
- UnipartiteGraph.py
- BipartiteGraph.py
- Graph.py
[1] Fabien Tarissan and Lionel Tabourier. A random model that relies on maximal bicliques to preserve the overlaps in bipartite networks. In 8th International Conference on Complex Networks and their Applications, Lisbonne, Portugal, December 2019.
[2] Émilie Coupechoux and Fabien Tarissan. Un modèle pour les graphes bipartis aléatoires avec redondance. In 4ème Journées Modèles et l’Analyse des Réseaux : Approches Mathématiques et Informatique (MARAMI’13), Saint-Etienne, France, October 2013.
[3] Peter Damaschke. Enumerating maximal bicliques in bipartite graphs with favorable degree sequences. Information Processing Letters, 114:317–321, June 2014.
[4] Enver Kayaaslan. On enumerating all maximal bicliques of bipartite graphs. pages 105–108, 01 2010.