How to Win Arguments

Weber, Klaus; Rach, Niklas; Minker, Wolfgang; André, Elisabeth

doi:10.1007/s13222-020-00345-9

Cited by 6 publications

(10 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Moreover, we will investigate if the machine learning performance can be improved if the user assumes a more active role in the interaction and include additional features that are accessible in such a case. In addition, we will look at techniques that can utilize the estimated information for adaptation, as for example Reinforcement Learning [39].…”

Section: Discussionmentioning

confidence: 99%

Estimating Subjective Argument Quality Aspects From Social Signals in Argumentative Dialogue Systems

et al. 2021

Self Cite

View full text Add to dashboard Cite

Information about a subjective user opinion towards an argument is crucial for argumentative systems in order to present appropriate content and adapt their behaviour to the individual user. However, requesting explicit feedback regarding the discussed arguments is often impractical and can hinder the interaction. To address this issue, we investigate the automatic recognition of user opinions towards arguments that are presented by means of a virtual avatar from social signals. We focus on two different user opinion categories (convincing and interesting) and two different types of social signals (facial expressions and eye movement). The recognition is addressed as a supervised learning problem and realized using the argument search evaluation data discussed in previous work. The overall performance is compared to a human annotation on a subset of the collected data. The results show that the machine learning performance is similar to human performance in both recognition tasks.

show abstract

Section: Discussionmentioning

confidence: 99%

Estimating Subjective Argument Quality Aspects From Social Signals in Argumentative Dialogue Systems

et al. 2021

Self Cite

View full text Add to dashboard Cite

show abstract

“…Within this work, we include both aspects into a separate agent strategy, which enables us to compare their persuasive effectiveness and investigate subliminal biases. The presented version of our system extends the previous one [42] in multiple ways. First, we utilize a modified version of the interaction model that is motivated by enabling more natural and intuitive persuasive dialogues between the virtual agents [26].…”

Section: Introductionmentioning

confidence: 90%

“…Each utterance is presented by the avatar of the corresponding agent with synthetic speech and emotion in the interface. As in the prototype version [42], the interface is based on the Charamel™ avatar 2 and utilizes Nuance TTS in combination with Amazon Polly Voices 3 . In addition to both avatars, the interface also includes buttons that enable users to assess the presented turn if it includes an argument.…”

Section: Systemmentioning

confidence: 99%

“…However, isolating and investigating the contribution of these individual aspects is difficult in human dialogues, as humans usually act and react intuitively in a conversation. To address this issue, we introduce a new version EVA 2.0 of the multi-agent system proposed in [42] in which different persuasion strategies can be displayed and compared. To this end, each agent is represented by a virtual avatar that interacts with its counterpart through synthetic voice and multimodal emotions.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

EVA 2.0: Emotional and rational multimodal argumentation between virtual agents

Rach

Weber

Yang

et al. 2021

It - Information Technology

Self Cite

View full text Add to dashboard Cite

Persuasive argumentation depends on multiple aspects, which include not only the content of the individual arguments, but also the way they are presented. The presentation of arguments is crucial – in particular in the context of dialogical argumentation. However, the effects of different discussion styles on the listener are hard to isolate in human dialogues. In order to demonstrate and investigate various styles of argumentation, we propose a multi-agent system in which different aspects of persuasion can be modelled and investigated separately. Our system utilizes argument structures extracted from text-based reviews for which a minimal bias of the user can be assumed. The persuasive dialogue is modelled as a dialogue game for argumentation that was motivated by the objective to enable both natural and flexible interactions between the agents. In order to support a comparison of factual against affective persuasion approaches, we implemented two fundamentally different strategies for both agents: The logical policy utilizes deep Reinforcement Learning in a multi-agent setup to optimize the strategy with respect to the game formalism and the available argument. In contrast, the emotional policy selects the next move in compliance with an agent emotion that is adapted to user feedback to persuade on an emotional level. The resulting interaction is presented to the user via virtual avatars and can be rated through an intuitive interface.

show abstract

“…The first argument search systems appeared recently [31,33], presenting both pros and cons instead of a direct answer [22], but still resembling the classical search engine interface. To provide for an interaction closer to human discussions, related research suggests employing either voice search [16] or avatars [35]. Even beyond, IBM's Project Debater [27] is the first system able to engage in a formal debate with humans.…”

Section: Conversational Argument Searchmentioning

confidence: 99%

The Meant, the Said, and the Understood: Conversational Argument Search and Cognitive Biases

Kiesel

Spina

Wachsmuth

et al. 2021

CUI 2021 - 3rd Conference on Conversational User Interfaces

View full text Add to dashboard Cite

Many questions of public interest do not have a single answer but come with a set of choices, each of which with its pros and cons. An "objective" information system can help explore the underlying argument space, and, if equipped with a conversational interface, it can create the experience of lively discussions resembling those from our daily lives. However, users will (subconsciously) extend the provided information by assumptions that adhere to their cognitive biases. In this regard, note that biases do not arise only from the underlying data or the employed algorithms, but also from the way the information is presented-especially in audio-only channels. Our paper brings attention to bias-related challenges of conversational interfaces for argument search systems. We identify research questions that address these challenges, and we propose ideas and methods to tackle them. CCS CONCEPTS• Human-centered computing → Interaction paradigms; • Information systems → Users and interactive retrieval.

show abstract

How to Win Arguments

Cited by 6 publications

References 41 publications

Estimating Subjective Argument Quality Aspects From Social Signals in Argumentative Dialogue Systems

Estimating Subjective Argument Quality Aspects From Social Signals in Argumentative Dialogue Systems

EVA 2.0: Emotional and rational multimodal argumentation between virtual agents

The Meant, the Said, and the Understood: Conversational Argument Search and Cognitive Biases

Contact Info

Product

Resources

About