| Guillaume Bouchard is a machine learning researcher focusing on statistical methods for reasoning on top of large scale knowledge bases. He is currently senior scientist at Xerox Reseach Centre Europe (XRCE), working on automatic dialogue systems to enable natural language interaction with a virtual agent to help customer find solutions to their problems.
He obtained a PhD in statistics from Institut National de Recherche en Information et Automatique (INRIA) in 2004. Since then, he worked on multiple machine learning research projects in big data analysis, including user modelling, recommender systems and natural language processing. His current interest is in distributed statistical relational learning with applications to dialogue sytems. He was involved as WP leader in French and European research projects called FUPOL, Fusepool and Dynamicité.
- Automated Agent: create an end-to-end solution for a virtual costumer agent. This involves a lot of statistical techniques to understand what the costumer says (Natural Language Understanding), what he/she wants to to (Predictive modelling) and how to solve it (Reinforcement Learning for Dialogues)
- City mobility analytics: Model how citizen use public transport. Micro-simulation of transportation system.
- Device log mining: analysing data coming from devices
- Learning analytics: predict student performances based on the teacher and the school
- FUPOL: EU-funded project in the e-governement domain. Looks at social media analytics to support policy makers.
- Fusepool : A content management system with a seamless integration of prediction engines. You can use machine learning algorithms without noticing it!
- Dynamicité: A social-media analytics project dedicated to the analysis of city data. In collaboration with Xerox, UTC and Linkfluence.
Students in 2014
- Théo Trouillon (2014-): PhD student with Eric Gaussier (UJF). Knowledge base factorization
- Chunyang Xiao (2014-): PhD student with Marc Dymetman (XRCE). Semantic parsing for dialogue
- Ehsan Abbasnejad: Intern from NICTA. Approximating logic using distributed representations
- Xiaofei Zhao: Inern from University of Montreal. Distributed learning of tensors.
- Bryan Feney: PhD student with Mark Girolami (UCL, UK): topic modeling
- All my past students
- Matrix Factorization
- Distributed representations, Knowledge Base Factorization and Tensor modeling
- Convex Optimization, Proximal Methods and Structured Sparsity
- Natural Language Processing (NLP)
- Variational Inference & High-dimensional integration
- Privacy-Preserving Distributed Machine Learning
- Generative-Discriminative Learning
More info on this page
- 2013: Big data mini-symposium. I organized a big data workshop at the SMAI (French SIAM
organization) where Zaid Harchaoui, Antoine Bordes, Jean-Philippe Vert
and Yann Stadnicki (Criteo) where invited to give a talk.
- 2011: Choice Models and Preference Learning workshop at NIPS with Shengbo Guo, Scott Sanner, Martin Szummer, Kristian Kersting, Paollo Viappiani, Onno Zoeter, Cedric Archambeau and Jean-Marc Andreoli.
- 2010:Automated Knowledge-Base Construction (AKBC) workshop in Grenoble
, jointly organized with Andrew McCallum, Cedric Archambeau, Onno Zoeter and Jean-Marc Andreoli
- 2009: NIPS Workshop on the Generative and Discriminative Learning Interface,
jointly organized with Percy Liang and Simon Lacoste-Julien;
- 2008: UAI/ICML workshop on Prior Knowledge for Text and Language Processing
jointly with Hal Daumé III, Marc Dymetman and Yee Whye Teh.