Skip to content



Intelligent systems are increasingly envisioned to support humans in diverse social settings, with examples ranging from intelligent tutoring applications to social robots in elderly care. Essential components for such systems to be successful are their abilities to (1) sufficiently estimate how users think and feel during interactions, in addition to (2) displaying appropriate behavior in response to these estimates. While Affective Computing (AC) research has tackled many technical aspects of these challenges, it has thus far left a fundamental barrier to progress largely unaddressed: the highly context-sensitive nature of human cognitive-affective processing. For example, how we express our feelings in terms of behavioral signals may depend strongly on the social setting in which we are currently embedded (e.g., the behavior of others present and the social norms and values these adhere to).

With this workshop, we aim to provide an interdisciplinary platform for discussing research on considering Social Context in Affective Computing systems. We strive to provide a platform to stimulate joint research projects, exchange methods, and a critical discussion of current and future efforts.

The ASOCA workshop is a satellite event of the 12th International Conference on Affective Computing and Intelligent Interaction (ACII 2023) taking place on 10 September 2023 at the MIT Media Lab, Cambridge, MA (USA).

  • Context-sensitive Modelling of cognitive-affective states in social situations from behavioral (e.g., facial expressions or gestures) and physiological signals (e.g., EEG, EDA, EMG, HR)..
  • Modeling internal user states during social interactions (e.g., emotions in group settings).
  • Personalization and Context-sensitivity in Human-robot (e.g., social robotics) and Human-computer interaction.
  • Personalization and context-sensitivity in social interactions with intelligent systems (e.g., conversational interactions with robots or virtual agents)
  • Context-sensitive adaptation to cognitive-affective user states (e.g., detecting and integrating features of social context).
  • Multi-modal datasets for modeling emotional and cognitive processes in social contexts (especially corpora spanning multiple different contextual settings).
  • Simulations of contextual influences on cognitive-affective processing or behavior in social interactions.

Important Dates

All deadlines are at the end of the day in the GMT-12 timezone.

Submission Deadline: 12 May, 2023 (extended)
Acceptance Notifications: 9 June, 2023 (extended)
Camera-ready Deadline: 1 August, 2023
Workshop date: 10 September, 2023


We invite submissions of the following types for presentation at the workshop:

  • Full Papers: max. 8 pages (7 pages + 1 page for references)
  • Short Papers: max. 5 pages (4 pages + 1 page for references)

Submissions should be double-blind, i.e., anonymous, and should follow the official submission guidelines from ACII2023.

Each paper will be sent to at least two expert reviewers and will have one of the organizers assigned as editor.

Accepted submissions will be made available on the workshop proceedings of ACII 2023. At least one author must register for the workshop and one conference day. Furthermore, we plan a special edition for the IEEE Transactions on Affective Computing (TAFFC) in the future on the workshop topic for extended papers.

Papers can be submitted via ACII’s EasyChair platform (choose track “Workshop: Addressing Social Context in Affective Computing”).


Time slots will correspond to the local time in Boston, Massachusetts (GMT-4)

Location: MIT Campus, Cambridge, MA, USA

09:00 – 09:10Welcome Note
09:10 – 09:55Invited Talk 1
Daniel Balliet: Perceptions of (social) situations explain variation in behavior
10:00 – 10:30Coffee Break
10:30 – 10:50Paper Blitz Talk (Session 1)
10:30 – 10:35
Vered Aharonson, Aviad Malachi and Tal Katz-Navon: Affective learning and the charismatic lecturer
10:35 – 10:40
Aarti Malhotra, Garima Sharma, Rishikesh Kumar, Abhinav Dhall and Jesse Hoey: Social Event Context and Affect Prediction in Group Videos
10:40 – 10:45
Bronagh Allison and Gary Mckeown: Are facial expression technologies tools for social interaction analysis rather than for
emotion recognition?
10:45 – 10:50
Mohammad Hasan Rahmani, Rafael Berkvens and Maarten Weyn: ColEmo: A Flexible Open Source Software Interface for Collecting Emotion Data
10:50 – 11:20 Poster Session 1*
*Presenting papers from the Blitz Talk Session 1
11:20 – 12:00Brainstorming Session in Groups (Part 1)

12:00 – 13:30

Lunch Break

13:30 – 14:15Invited Talk 2
Giovanna Varni: Which roles does affect play in social contexts?
14:20 – 14:35Paper Blitz Talk (Session 2)
14:20 – 14:25
Qiurui Chen, Laduona Dai and Nino Hardt: Comparative Analysis of Vocal and Textual Emotion Detection and their association
with Consumer Preferences: An Empirical Study
14:25 – 14:30
Yajing Feng and Laurence Devillers: End-to-end Continuous Speech Emotion Recognition in Real-life Customer Service
Call Center Conversations
14:30 – 14:35
Pranavi Jalapati, Selwa Sweidan, Xin Zhu and Heather Culbertson: Vocalization for Emotional Communication in Crossmodal Affective Display
14:40 – 15:15Poster Session 2*
*Presenting papers from the Blitz Talk Session 2
15:15 – 15:30 Coffee Break
15:30 – 16:10Brainstorming Session in Groups (Part 2)
16:10 – 16:40 Plenary Discussion
16:40 – 16:50 Closing

Invited Speakers

Giovanna Varni

Talk title: Which roles does affect play in social contexts?

Talk abstract: The role played by affect in social context, although a major goal of affective computing, has received less attention compared to modeling and synthesizing individual affect. In my talk I will present my studies on affect in dyadic and team social context to show how affect can be either a shaper or a by-product of interaction.

Bio: Giovanna Varni is an Associate Professor at Department of Information Engineering and Computer Science (DISI), University of Trento, Italy. She is an interdisciplinary researcher mainly investigating on Social Signal Processing (SSP), Affective Computing (AC) and Human Computer Interaction (HCI). She was involved in several EU FP7-FP6 projects, and she was PI of the national French project ANR JCJC GRACE (2019-2022) on the automated analysis of cohesion in small groups of humans. She contributes regularly to organizational roles in international conferences and workshops relevant for her specific research area such as ACII and ICMI, for which she also serves as a Program Committee member.

Talk title: Perceptions of (social) situations explain variation in behavior

Talk abstract: Humans must understand the context of their decisions and behavior to behave in ways that are best for themselves and others. Recently, there has been an abundance of models about how humans understand situations they experience in daily life and its importance for behavior. I will first briefly review a few of the models, and then center my attention on a specific theory of how people understand social situations, that is Functional Interdependence Theory (FIT). FIT proposes that people experience a great variety of interdependent situations with others and that there could exist benefits from detecting the type of interdependence people experience in a situation and using that to condition behavioral strategies. Interdependence can vary along four dimensions, and I will discuss evidence that people can perceive situations (and relationships) along these dimensions. Across several studies, we have found that people can reliably differentiate situations along five dimensions (mutual dependence, corresponding-versus-conflicting outcomes (i.e., conflict), asymmetry of dependence (i.e., Power), future interdependence, and information certainty). Moreover, people can use these dimensions to describe different relationships in their social network (e.g., acquaintances, friends, romantic partners, and family). Across several studies, we have observed how people perceive their social interactions and predict when they decide to cooperate. An ability to perceive differences in interdependent situations and relationships could enable people to make better partner choices, condition behavioral strategies, and adapt to a broad range of ecological conditions which vary according to interdependence.

Bio: Daniel Balliet’s research focuses on understanding Human Cooperation. He applies experiments, field studies, and meta-analysis to test evolutionary and psychological theories of cooperation. His work addresses issues related to (a) how people think about their interdependence in social interactions, (b) how people condition their cooperation to acquire direct and indirect benefits, and (c) understanding cross-societal variation in cooperation. He is the recipient of an ERC Starting Grant (2015-2020) and an ERC Consolidator Grant (2020-2025).

Organizing Committee

Bernd Dudzik
Delft University of Technology

Tiffany Matej Hrkalovic
VU Amsterdam & Delft University of Technology

Joost Broekens
Leiden University

Dirk Heylen
University of Twente

Zakia Hammal
Carnegie Mellon University

For inquiries, please contact us via

Program Committee