Towards a Model of Face-to-Face Grounding.
The Role of Lexico-Semantic Feedback in Open-Domain Textual Question-Answering.