Automatic evaluation of conversation qualities has become one of the vital points for interactive computer systems (ICCs) of the next generation. Although current ICCs are able to understand what humans are talking about, they are incapable of catching how humans are talking. Moreover, recent research on human-human interaction has shown that more than 90 % of the information contained in speech and visual input is non-verbal. Important parts of this information include conversational qualities (characteristics). While implicitly transmitted among humans during the conversation, they significantly influence the entire conversation and therefore its quality. Modern robotic and computer systems may not even suspect that something is going wrong during the interaction until it will be “accidentally” aborted. However, if such systems were capable to catch and to analyse all available conversational qualities shown by human agents, they would act appropriately, embodying naturalness, confidence, and intelligibility.

When people talk to each other, they change their verbal and nonverbal communication behaviors according to those of the partner. Therefore, user adaptivity represents an essential issue in improving the quality of human-robot and, more generally, human-agent interaction.There are many potential applications where user adaptivity adds value for enhancing user experience. For example, E-Health is quite relevant as a use case for adaptive human-computer interactions with AI systems since tailored and explainable interventions are needed for long term engagement. A virtual coach needs to know about the user's personality and cultural background since they impact the user's consideration of health and motivation.

The aim of this Shonan meeting is therefore to discuss challenges that may arise in Conversational Qualities Assessment (CQA) during human-human and human-robot/agent interaction. Possible usage of these CQA Systems in various industrial spheres with focus on application within Human-Computer (Robot) interactions will also be discussed. Finally, a roadmap of the technology development will be established.

To move closer to the solutions of the challenges described above, significant efforts are required from the scientific community involving specialists in the field of computer science, medicine, and psychology. Sharing and combining expertise from various scientific fields may lead to synergies, which will allow us to create new ways to build solutions or increase the effectiveness of existing approaches. The Shonan Meeting will help to explore possible challenges and jointly develop the respective research fields, whose contributions will serve as a research agenda for main directions. Therefore, we will invite keynote speakers from the respective research fields, whose contributions will serve as a basis for breakout sessions. In these sessions, participants will work actively on specific research objectives in small groups. This will help foster an interdisciplinary understanding and cooperativity. The results of the breakout sessions will then be discussed with the whole plenum.

We have decided to structure the workshop into four different areas.

• Multimodal CQA in Human-Human Interaction and HRI
  • Influence of paralinguistics on CQA.
  • Multimodal Machine Learning for modelling dyadic and group interaction.
  • Multimodal CQA: approaches and fusion techniques.
  • Online versus face-to-face communication.
• Human-Robot and Human-Human Interaction
  • Human-Computer and Human-Robot Interaction making use of CQA.
  • CQA of Human-robot interaction strategies in terms of adaptivity.
  • Group level performance modelling (Cohesion, Group output).
  • CQA of diverse groups (human-only group, human-robot group).
• Influence of User Properties, Personality and Emotions
  • Personality trait modelling (Personality, Attitude, Engagement, Social skills).
  • Modelling of influence of individual differences (age, gender, and language) on multimodal interaction.
  • Affective modelling on multimodal interaction (Sentiment, Empathy).
  • The role of emotions in human-robot conversations.
• Resources and Data for CQA
  • Data collection: setup and choice of sensors.
  • Available Tools for data collection, processing, and annotation
  • Data robustness: suppressing noise, learning from small data and handling missing data.
  • Data quality: Reliable annotation and motivation of participants.

• Experimental design, user studies, and evaluation of systems for automatic CQA.
• Experimental set-up accounting for real-world, real-time, large-scale conditions.
• Engineering approaches to CQA: life-cycle, requirement elicitation, robustness to change, standardization and simulation.
• Development models, tools and strategies: middleware, languages.
• Meaningful and explainable system evaluation.
• Description, development and sharing of resources: corpora compilation, annotation tools and approaches, crowdsourcing approaches.
• Sensing devices and frameworks designed for CQA.
• Competitive research challenges planning and organisation.

• Computer-mediated human-to-human interaction.
• Health applications
• Social Human-Robot Interaction.
• Social skill training applications
• Tutoring and E-learning applications.
• Coaching applications.
• Energy, memory, and computing efficient CQA. Model pruning/shrinking for usage on portative devices.
• Success stories, functional systems and industrial challenges.

• Data protection and privacy by design and default.
• Legal issues.
• Trust and usability.
• Social responsibility.
• Interdisciplinary approach by integrating findings of physiology and social science

The outcome of the workshop will be published in the form of free open-access (http://ceur-ws.org/). This is expected to encourage joint publications at top conferences and journals in computer science, jointly authored by psychologists, ethicists, and AI researchers.