Section: Bilateral Contracts and Grants with Industry

Bilateral Contracts with Industry

First, we are involved in the Hermes project along a collaboration with the Sequel inria team and with a consortium of companies. In that collaboration, the envisioned applications is the design of recommender systems for commercial data. One objective is to provide social recommendations, that is to take into accounts in the recommendations, social relationships between users and the content of messages posted by users in forums.

Second, we start a one to one cooperation with the Clic and Walk company along the PhD thesis of Pauline Wauquier . The company makes marketing surveys by consumers (called clicwalkers). The goal of the company is to understand the community of clicwalkers (40 thousands in one year) and its evolution with two objectives: the first one is to optimize the attribution of surveys to clicwalkers, and the second is to expand company's market to foreign countries. Social data can be obtained from social networks (G+, Facebook, ...) but there is no explicit network to describe the clicwalkers community. But users activity in answering surveys as well as server logs can provide traces of information diffusion, geolocalisation data, temporal data, sponsorship,... We will study the problem of adaptive graph construction from the clicwalkers network. Node (users) classification and clustering algorithms will be applied. For the problem of survey recommendations, the problem of teams constitution in a bipartite graphs of users and surveys will be studied. Random graph modeling and generative models of random graphs will be one step towards the prediction of the evolution of clicwalkers community.

Third, we have started a transfer collaboration with the Music Story company. In a first phase, we have considered the question of collecting musical metadata from heterogeneous sources. We have proposed machine learning methods and similarity measures for curating metadata. The Music Story company has close industrial collaborations with the Deezer company. Current discussions between Magnet and these two companies are open on social recommender systems for music.

Last, we work with physicians at the Lille hospital (CHRU) on the detection of brain anomalies related to epilepsy. Hence, we will use connectome data which is an approximate map of neural connections at different scales. The connectome can be modeled by a weighted graph. Available data include graphs constructed at different times for a given patient, also graphs for healthy patients and epileptic patients. One objective of the research project is to study how the connectome together with other signals, like functional magnetic resonance imaging (FRMI), MEG and EEG can be efficiently combined in order to detect abnormal brain regions. We will consider diffusion algorithms in graphs to test whether diffusion processes in the brain can be explained with the connectome. We will also consider learning algorithms related to information diffusion in order to enhance graph construction.