A User Simulator Architecture for Socially-Aware Conversational Agents

Jain, Alankar; Pecune, Florian; Matsuyama, Yoichi; Cassell, Justine

doi:10.1145/3267851.3267916

Cited by 17 publications

(13 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…R17 suggested it would be valuable to have ways to prototype conversational agents more rapidly, including methods for simulating conversational trajectories (cf. [56,113]) "and then find[ing] ways to automate the identification of risky conversation patterns that emerge. " Similarly, a data scientist working on adaptive mentoring software (R10) suggested that because their product involves a long-term feedback loop, it would be ideal to run it on a population of "simulated mentees" (cf.…”

Section: Needs For More Holistic Auditing Methodsmentioning

confidence: 99%

Improving Fairness in Machine Learning Systems

Holstein

Vaughan

Daumé

et al. 2019

Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems

571

449

View full text Add to dashboard Cite

The potential for machine learning (ML) systems to amplify social inequities and unfairness is receiving increasing popular and academic attention. A surge of recent work has focused on the development of algorithmic tools to assess and mitigate such unfairness. If these tools are to have a positive impact on industry practice, however, it is crucial that their design be informed by an understanding of realworld needs. Through 35 semi-structured interviews and an anonymous survey of 267 ML practitioners, we conduct the first systematic investigation of commercial product teams' challenges and needs for support in developing fairer ML systems. We identify areas of alignment and disconnect between the challenges faced by teams in practice and the solutions proposed in the fair ML research literature. Based on these findings, we highlight directions for future ML and HCI research that will better address practitioners' needs.

show abstract

Section: Needs For More Holistic Auditing Methodsmentioning

confidence: 99%

Improving Fairness in Machine Learning Systems

Holstein

Vaughan

Daumé

et al. 2019

Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems

571

449

View full text Add to dashboard Cite

show abstract

“…Rapport being a multi-party phenomenon, it depends on the behavior of all conversation participants. Hence, to optimize our agent's social behavior, we followed the approach described in [11] by building a user simulator in which actions are a combination of a task action and a conversational strategy. The task actions simulated users can take are listed in Table 1.…”

Section: Simulated Usermentioning

confidence: 99%

“…I-Type users, on the other hand, express a wide variety of conversational strategies, actively pursuing both task and social goals during their interaction. We learned from data [11] that 64.1% of the users are P-Type and the remaining 35.9% are I-Type. 1 https://www.imdb.com/ Our user simulator is agenda-based [21] and largely composed of handcrafted rules apart from one stochastic decision point learnt from data: recommendation acceptance.…”

Section: Simulated Usermentioning

confidence: 99%

A framework to co-optimize task and social dialogue policies using Reinforcement Learning

Pecune

Marsella

2020

Proceedings of the 20th ACM International Conference on Intelligent Virtual Agents

Self Cite

View full text Add to dashboard Cite

One of the main challenges for conversational agents is to select the optimal dialogue policy based on the state of the interaction. This challenge becomes even harder when the conversational agent not only has to achieve a specific task, but also aims at building rapport. Although some work already tried to tackle this challenge using a Reinforcement Learning (RL) approach, they tend to consider one single optimal policy for all the users, regardless of their conversational goals. In this work, we describe a framework that allows us to build a RL-based agent able to adapt its dialogue policy depending on its user's conversational goals. After we build a rulebased agent and a user simulator communicating at the dialog-act level, we crowdsource the surface sentences authoring for both the simulated users and the agent, which allow us to generate a dataset of interactions in natural language. Then, we annotate each of these interactions with a single rapport score and analyze the links between simulated users' conversational goals, agent conversational policies, and rapport. Our results show that rapport was higher when both or none of the interlocutors tried to build rapport. We use this result to inform the design of a social reward function, and we rely on this social reward function to train a RLbased agent using an hybrid approach of supervised learning and reinforcement learning. We evaluate our approach by comparing two different versions of our RL-based agent: one that takes users' conversational goals into account and another that does not. The results show that an agent adapting its dialogue policy depending on users' conversational goals performs better.

show abstract

“…This can be approached for example as an optimisation or ranking problem using deep learning methods as in [22]. Other possibilities include adopting an adaptive dialogue policy based on a social model of the user and the current status of their relationship with the agent [8].…”

Section: Discussionmentioning

confidence: 99%

It's Good to Chat?

Gunson

Sieińska

Walsh

et al. 2020

Proceedings of the 20th ACM International Conference on Intelligent Virtual Agents

View full text Add to dashboard Cite

We present and evaluate a deployed conversational AI system that acts as a host of a working public building on a university campus. The system combines open-domain social chat with task-based conversation regarding navigation in the building, live resource updates (e.g. available computers), and events in the building. We investigated the impact of open-domain social chat on task completion and user preferences by comparing the combined system with a task-only version. We find that there is no significant difference in task completion or several aspects of user preference between the two systems, but that users would be significantly happier to talk to the task-only system in the future. This suggests that the łwalk-upž public setting and workplace nature of the environment creates a markedly different use case to the in-home, and more individual and private łcompanion/assistantž setting which is commonly assumed for systems like Alexa. We discuss the implications for the design of conversational systems in other public settings. CCS CONCEPTS • Human-centered computing → Natural language interfaces; Empirical studies in HCI; • Computing methodologies → Discourse, dialogue and pragmatics.

show abstract

A User Simulator Architecture for Socially-Aware Conversational Agents

Cited by 17 publications

References 30 publications

Improving Fairness in Machine Learning Systems

Improving Fairness in Machine Learning Systems

A framework to co-optimize task and social dialogue policies using Reinforcement Learning

It's Good to Chat?

Contact Info

Product

Resources

About