ERC Consolidator Grant awarded to Raquel Fernández
We are pleased to announce that Raquel Fernández has been awarded an ERC Consolidator, to work on Distributed dynamic representations for dialogue management (DREAM).
Our ability to communicate using language in conversation is considered the hallmark of human intelligence. Yet, while holding a dialogue is effortless for most of us, modelling this basic human skill by computational means has proven extremely difficult. Fernández will address this challenge by establishing a new computational model of a dialogue agent that can learn to take part in conversation directly from data about language use. Her model is grounded in linguistic theories of dialogue, but exploits recent advances in computational learning that allow the agent to learn the representations that it manipulates directly from experience. This constitutes a paradigm shift in dialogue modelling --- from predefined symbolic representations to automatic representation learning – that will break new scientific ground in (computational) linguistics, and artificial intelligence. Fernández will implement the agent as an artificial neural network system and train it with task-oriented conversations where the participants have a well-defined end goal. The agent will be able to integrate linguistic and perceptual information, leading to more human-like and effective communication.