Pragmatic Frames for Teaching and Learning in Human–Robot Interaction: Review and Challenges

Vollmer, Anna-Lisa; Wrede, Britta; Rohlfing, Katharina J.; Oudeyer, Pierre-Yves

doi:10.3389/fnbot.2016.00010

Cited by 23 publications

(17 citation statements)

References 63 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Most of existing advice-taking systems assume that the internal representation of advice is predetermined by the system designer. However, some recent works tackle the problem of letting the system learn how to interpret raw advice in order to make the interaction protocol less constraining for the human teacher (Vollmer et al, 2016 ). Steps 3–5 describe how human advice can be used by the agent for learning.…”

Section: Reinforcement Learning With Human Advicementioning

confidence: 99%

“…Since then, many efforts have been made for endowing robots and artificial agents with the capacity to learn from humans in a natural and unconstrained manner (Chernova and Thomaz, 2014 ). However, designing human-like learning robots still raises several challenges regarding their capacity to adapt to different teaching strategies and their ability to take advantage of the variety of teaching signals that can be produced by humans (Vollmer et al, 2016 ).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Reinforcement Learning With Human Advice: A Survey

Najar

Chétouani

2021

Front. Robot. AI

View full text Add to dashboard Cite

In this paper, we provide an overview of the existing methods for integrating human advice into a reinforcement learning process. We first propose a taxonomy of the different forms of advice that can be provided to a learning agent. We then describe the methods that can be used for interpreting advice when its meaning is not determined beforehand. Finally, we review different approaches for integrating advice into the learning process.

show abstract

Section: Reinforcement Learning With Human Advicementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Reinforcement Learning With Human Advice: A Survey

Najar

Chétouani

2021

Front. Robot. AI

View full text Add to dashboard Cite

show abstract

“…The significance of this study is the need for a renewed global initiative in education to promote debates, research and on-going collaboration with scientific leaders on ethics and programming robots. The implication for education leaders is to provide ongoing professional development on the role of ethics in education and to create best practices for using robots in education to promote increased student learning and enhance the teaching process (Vollmer et al, 2016).…”

Section: Findings Summarymentioning

confidence: 99%

Perceptions of robotics emulation of human ethics in educational settings: a content analysis

Fedock

Paladino

Bailey

et al. 2018

Jrnl of Rsrch Inno Teach & Lrn

View full text Add to dashboard Cite

Purpose-The purpose of this paper is to examine how robotics program developers perceived the role of emulation of human ethics when programming robots for use in educational settings. A purposive sampling of online robotics program developer professional sites which focused on the role of emulation of human ethics used when programming robots for use in educational settings was included in the study. Content related to robotics program developers' perceptions on educational uses of robots and ethics were analyzed. Design/methodology/approach-The design for this study was a qualitative summative content analysis. The researchers analyzed keywords related to a phenomenon. The phenomenon was the emulation of human ethics programmed in robots. Articles selected to be analyzed in this study were published by robotics program developers who focused on robots and ethics in the education. All articles analyzed in this study were posted online, and the public has complete access to the studies. Findings-Robotics program developers viewed the importance of situational human ethics interpretations and implementations. To facilitate flexibility, robotics program developers programmed robots to search computer-based ethics related research, frameworks and case studies. Robotics program developers acknowledged the importance of human ethics, but they felt more flexibility was needed in the role of how classroom human ethical models were created, developed and used. Some robotic program developers expressed questions and concerns about the implementations of flexible robot ethical accountability levels and behaviors in the educational setting. Robotics program developers argued that educational robots were not designed or programmed to emulate human ethics. Research limitations/implications-One limitation of the study was 32 online, public articles written by robotics program designers analyzed through qualitative content analysis to find themes and patterns. In qualitative content analysis studies, findings may not be as generalizable as in quantitative studies. Another limitation was only a limited number of articles written by robotics programs existed which addressed robotics and emulation of human ethics in the educational setting. Practical implications-The significance of this study is the need for a renewed global initiative in education to promote debates, research and ongoing collaboration with scientific leaders on ethics and programming robots. The implication for education leaders is to provide ongoing professional development on the role of ethics in education and to create best practices for using robots in education to promote increased student learning and enhance the teaching process. Social implications-The implications of this study are global. All cultures will be affected by the robotics' shift in how students are taught ethical decision making in the educational setting. Robotics program developers will create computational educational moral models which will replace archetypal educational ethics framewo...

show abstract

“…Because the builder imitates itself, we call it self-imitation. The notion of interaction frames (also called pragmatic frames) states that agents that interact in sequence can more easily interpret the interaction history (Bruner, 1985;Vollmer et al, 2016). In ABIG, we consider two distinct interaction frames.…”

Section: Introductionmentioning

confidence: 99%

Learning to Guide and to Be Guided in the Architect-Builder Problem

Barde¹,

Karch²,

Nowrouzezahrai³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

We are interested in interactive agents that learn to coordinate, namely, a builder -which performs actions but ignores the goal of the task -and an architect which guides the builder towards the goal of the task. We define and explore a formal setting where artificial agents are equipped with mechanisms that allow them to simultaneously learn a task while at the same time evolving a shared communication protocol. Ideally, such learning should only rely on high-level communication priors and be able to handle a large variety of tasks and meanings while deriving communication protocols that can be reused across tasks. The field of Experimental Semiotics has shown the extent of human proficiency at learning from a priori unknown instructions meanings. Therefore, we take inspiration from it and present the Architect-Builder Problem (ABP): an asymmetrical setting in which an architect must learn to guide a builder towards constructing a specific structure. The architect knows the target structure but cannot act in the environment and can only send arbitrary messages to the builder. The builder on the other hand can act in the environment but has no knowledge about the task at hand and must learn to solve it relying only on the messages sent by the architect. Crucially, the meaning of messages is initially not defined nor shared between the agents but must be negotiated throughout learning. Under these constraints, we propose Architect-Builder Iterated Guiding (ABIG), a solution to the Architect-Builder Problem where the architect leverages a learned model of the builder to guide it while the builder uses self-imitation learning to reinforce its guided behavior. To palliate to the non-stationarity induced by the two agents concurrently learning, ABIG structures the sequence of interactions between the agents into interaction frames. We analyze the key learning mechanisms of ABIG and test it in a 2-dimensional instantiation of the ABP where tasks involve grasping cubes, placing them at a given location, or building various shapes. In this environment, ABIG results in a low-level, high-frequency, guiding communication protocol that not only enables an architect-builder pair to solve the task at hand, but that can also generalize to unseen tasks. * Equal contribution.† Work conducted while at Inria.

show abstract

Pragmatic Frames for Teaching and Learning in Human–Robot Interaction: Review and Challenges

Cited by 23 publications

References 63 publications

Reinforcement Learning With Human Advice: A Survey

Reinforcement Learning With Human Advice: A Survey

Perceptions of robotics emulation of human ethics in educational settings: a content analysis

Learning to Guide and to Be Guided in the Architect-Builder Problem

Contact Info

Product

Resources

About