Generating Hints and Other Assistance

automated generation of hints to help user

Hint Engine

The Hint Engine is an AI algorithm to generate an appropriate hint for the user, enabling them to make progress on the task. The algorithm combines information from multiple models to make this decision (Wilson et al., 2019).
It uses a task model to identify the next action the user needs to take (Wilson et al., 2018) It uses an assistance model to map actions in a task to specific types of hints or assistance that the robot can provide. A model of the user’s level of need guides the algorithm in deciding how direct of a hint to provide. The underlying theory is that providing only as much help as is needed will help maintain the autonomy of the user. Lastly, a user preferences model captures two types of user preferences, which are used to guide the algorithm in understanding how the user wants to perform the task (Wilson et al., 2020).

References

2020

  1. A Knowledge Driven Approach to Adaptive Assistance Using Preference Reasoning and Explanation
    Jason R. Wilson, Leilani Gilpin , and Irina Rabkina
    In Proceedings of the AI-HRI Symposium at AAAI-FSS 2020 , 2020

2019

  1. Developing Computational Models of Social Assistance to Guide Socially Assistive Robots
    Jason R. Wilson, Seongsik Kim , Ulyana Kurylo , and 2 more authors
    In Proceedings of the AAAI Fall Symposium on Artificial Intelligence and Human-Robot Interaction for Service Robots in Human Environments , 2019

2018

  1. General Approach to Automatically Generating Need-based Assistance
    Jason R. Wilson, Ram Wransky , and Jorge Tierno
    In Proceedings of the Sixth Annual Conference on Advances in Cognitive Systems , 2018