I am currently working on Hypergraph modeling and Visualisation of Co-occurence Networks.

The aim is to gain insights into textual datasets.

Co-occurences are n-adic relationships and are well modeled by hb-graphs, an extension of hypergraphs, that I introduced.

Hb-graphs are families of sub-multisets of a vertex set. A multiset allows the repetition of elements; they are also called bags or msets.

Simplifying the approach one can retrieve hypergraphs, which are families of subsets of vertices.

To visualise different facets of an information space, we built a 2.5D layout that allows to switch between the different facets through the chosen reference and to show co-occurences built vs this reference inside the different facets.