Do Humans Adapt Their Emails So That Agents Can Understand?
نویسندگان
چکیده
Natural language processing is a difficult problem and is currently a technical barrier in building personal assistant agents that aim to interact with humans in natural language. One possible solution is to rely on humans to restrict or adapt their language into more computer friendly ways. We study the feasibility of this approach in the context of a personal assistant agent that parses emails relevant to meeting scheduling in order to assist the user in calendar management tasks. We design an experiment in which humans are given email writing tasks and then given feedback on how well those emails were understood by an agent. We wish to see if the humans learn to adapt their email writing to what is understandable by their agent.
منابع مشابه
Using Virtual Humans to Study Embodied Social Interaction
An important step to build machines that can learn from social interactions with human users is to understand the nature of learning-oriented interactions. To do so, a central problem is to find a way to decouple the social interaction between two agents (e.g. a human supervisor and a machine learner), so that we can systematically manipulate and control the dynamic flow of the interaction to c...
متن کاملLearning through Multimodal Interaction
This paper proposes and implements a new experimental paradigm to study the role of multimodal interaction in automatic language learning. We observe that child language acquisition relies significantly on everyday social interactions with adult partners. In light of this, we argue that an important step to build machines that can also learn from social interactions with human users is to under...
متن کاملTempo Adaptation and Anticipation Methods for Human-Robot Teams
As technology advances, autonomous robots are becoming more prominent in our daily lives, and they will be expected to effectively work with groups of humans. In order to be effective teammates, robots need to be able to understand human team dynamics as humans do. This understanding will help robots to recognize, anticipate, and adapt to human motion. There exist many group interaction scenari...
متن کاملExplanation of Agent's Professional Capabilities in Agriculture Jihad for Rural Development of Kermanshah Province
Introduction Rural planning is one of the most important activities for agents and planners of development. Agriculture Jihad aims to accomplish rural development purposes through codification of plans and activities to increase the level of awareness and professional skills of villagers. In developing countries, it is a problem that rural planners do not have enough technical and profession...
متن کاملAutomation of social networks with QA agents
This paper describes a model, ASKNEXT, for connecting agents using social networks for knowledge exchanges using email. It proposes a protocol and a mathematical model to understand how emails spread throughout social networks of agents and people, which can be used to predict the scalability of agents exchanging emails to find answers to questions.
متن کامل