#oreilly2016

Gaze in Human-Robot Communication is a volume collecting recent research studying gaze behaviour in human-robot interaction (HRI)

Abstract Conversational modeling is an important task in natural language understanding and machine intelligence. Although previous approaches exist, they are often restricted to specific domains (e.g., booking an airline ticket) and require handcrafted rules. In this paper, we present a simple approach for this task which uses the recently proposed sequence to sequence framework. Our model converses by predicting the next sentence given the previous sentence or sentences in a conversation. The strength of our model is that it can be trained end-to-end and thus requires much fewer hand-crafted rules. We find that this straightforward model can generate simple conversations given a large conversational training dataset. Our preliminary suggest that, despite optimizing the wrong objective function, the model is able to extract knowledge from both a domain specific dataset, and from a large, noisy, and general domain dataset of movie subtitles. On a domain-specific IT helpdesk dataset, the model can find a solution to a technical problem via conversations. On a noisy open-domain movie transcript dataset, the model can perform simple forms of common sense reasoning. As expected, we also find that the lack of consistency is a common failure mode of our model.

Focus shifting in the repetitions occurred, but primarily among the adult users of our system. Children, on the other hand, tended not to increase the duration of their repetitions, but rather to speak louder. This difference in adult and children strategies could have a number of explanations. It could be argued that while adults believe the system did not ‘understand’ them the first time, children think the system did not ‘hear’ what they were saying.

DESIGN IMPLICATIONS Detecting whether or not people greet a computational agent provides an opportunity to design adaptive dialogue systems for cooperation. Social agents might use relational strategies with those who greet the agent and more utilitarian dialogue with those who do not greet the agent. People who spontaneously greet agents might be likely to respond more positively to agents that attempt small talk than agents that do not. Bickmore and Cassell showed that agents that made small talk reduced the perceived distance between themselves and users, and increased users’ trust, especially when users were extroverted