Postdoc for approximate graph mining

Postdoc for the ApproxiM project: approximate graph mining in large databases and streams at LIG lab

Duration: 12 months
Expected start date: Between June and October 2011

Context.
Nowadays, a huge quantity of useful information is represented in graphs. These graphs can appear in social networks such as Facebook or Twitter, in biology with gene interaction networks and also in the call logs of telecommunication operators.

The data mining technique called frequent subgraph mining allows one to analyze such graph data and extract subgraphs that occur frequently in the input graph data.

The problem is that frequent subgraph mining is an extremely heavy task, requiring long computation time, and it can hardly scale up to huge graphs, especially those found in social networks.

The goal of the postdoc in the ApproxiM project is to design approximate graph mining methods, which could scale to huge quantity of input data while providing subgraphs of sufficient quality.

This work will be applied to the analysis of diffusion of information in social networks.

Team: The candidate will work with Eric Gaussier and Alexandre Termier.

Prerequisites: The candidate is expected to have a strong background in computer science (graphs, algorithms,...) in general and in data mining in particular.
Good programming skills are also requested.

_Contact_:
Please mail your CV to both Alexandre.Termier@imag.fr and Eric.Gaussier@imag.fr