Connectionists: XRCE internship: Deep Recurrent Neural Networks for Dialogue Prediction and Response Generation

[Dymetman, Marc]

Xerox Research Centre Europe, located in Grenoble, France, is offering the following internship for early 2016:



--------------------------------------------------------------------------------------------------------------

Deep Recurrent Neural Networks for Dialogue Prediction and Response Generation

---------------------------------------------------------------------------------------------------------------



See: http://www.xrce.xerox.com/About-XRCE/Internships/Deep-Recurrent-Neural-Networks-for-Dialogue-Prediction-and-Response-Generation



Contact :  Marc Dymetman   marc.dymetman at xrce.xerox.com<mailto:marc.dymetman at xrce.xerox.com>



Please mention "RNNs for Dialogue and NLG" in your subject line.


Duration: around 5 months
Start date: around January/February 2016

The application of Deep Learning techniques to conversational interactions has recently emerged as an important research topic in the Machine Learning community, and XRCE has started to be strongly engaged in this area.

In particular, we are actively working in the domain of task-oriented dialogues in relation to the customer call-centers Xerox manages, and have access to significant amounts of interaction data in these domains.

We are looking for an intern who will help us improve some of our current Recurrent Neural Network models for dialogue prediction and generation for these task-oriented conversations.

The ideal candidate is Computer Science student at the Master or (preferably) PhD level, with a background in Machine Learning (preferably including Deep Learning techniques) and NLP. Strong programming skills are a requirement (Python, C++, Java...) and knowledge of such toolkits as Theano, Torch, TensorFlow would be a plus.

Conference publication of the work will be strongly encouraged.
See XRCE's page above for more details.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/connectionists/attachments/20151124/47d05a00/attachment.html>