Loading…
Registration open daily from 8am - 6pm.  Please join us for the #ISSS2015 #Roundtable at 7.45am each morning.
Back To Schedule
Thursday, August 6 • 16:30 - 17:00
Positeams – Positive Systems Intelligent Teams, an Agent-Based Simulator for Studying Group Behaviour

Sign up or log in to save this to your schedule, view media, leave feedback and see who's attending!

Systems intelligence is the ability to act intelligently within complex systems involving interactions and feedback. Social groups are systems, i.e. wholes consisting of multiple mutually interacting parts, where the interactions often include non-linearities and feedback loops. Agent-based simulations have been used extensively to model social systems. In this paper we use a new systems intelligence perspective when analysing the performance of a team. We consider and model the dynamics of positivity in a social system. The goal is to help understand the full consequences of our behaviour, both positive and negative, and how we could change it for the benefit of the whole system. Agent-based simulations of social networks can provide useful insights into the underlying systems and introduce ideas to improve their performance. The dynamics of social systems can be difficult to understand because of their systemic nature. In addition, one seldom has an opportunity to view social systems from the outside, but rather one is an active part within the system with some power to affect its state. Agent-based simulations support the systems approach by emphasizing the holistic view on understanding the problem rather than focusing on its parts in isolation, facilitating personal decision making in social systems.

PoSITeams is a web-based multi-agent social simulator that simulates the dynamics and evolution of positive and negative affect in a team. People are greatly influenced by the emotions of others and positive emotional contagion has been linked to increased performance in social groups. In particular, the ratio of positive and negative affect has proven to be an especially useful parameter, and it has been successfully applied to predicting effective organizations and successful marriages. Also on the individual level the characteristic difference between flourishing and non-flourishing individuals has been observed to be the ratio of experienced positive and negative emotions. Thus simulating emotion dynamics can be used to better understand how to act more constructively within organizations. Our model is a discrete-time non-linear dynamic model, where the positivity ratio defines the agent's interaction style. The level of emotional contagion depends on the strength of the social contact between the agents and their personal characteristics, such as tendency to express and be influenced by emotions. Furthermore, the positivity ratio of the agents increases their resiliency against negative emotions as indicated by the positivity research.

The purpose of the simulator is to enable the user to simulate social groups of her own and explore the effects of different behavioural and structural changes. We are especially interested in engaging the user in reflective thought-processes and facilitate seeing the system as a whole and let the user recognize herself as an active part of the system, who is both able to affect the state of the system and is reciprocally influenced herself by the system, a concept known as systems intelligence introduced by Esa Saarinen and Raimo P. Hämäläinen. The simulator is also able to suggest systems intelligent actions by optimizing the total positivity of the group.

Compared to the related agent-based models for emotional contagion, our approach emphasizes the qualitative differences between the negative and positive affect as described by psychological research, namely the broaden-and-build theory of positive emotions and the negativity bias, which motivate modelling the positive and negative affect separately. We also highlight the significance of positivity ratios to describe the performance of social groups.


Presenter / Artist
avatar for Teemu Tiinanen

Teemu Tiinanen

Aalto University
I currently work on systems intelligence, which is the ability to act intelligently within complex systems involving interaction and feedback. (see http://systemsintelligence.aalto.fi/). I am especially interested in developing methods to support systems intelligent behaviour in organizations.I... Read More →


Thursday August 6, 2015 16:30 - 17:00 CEST
Copenhagen 2 Scandic Berlin Potsdamer Platz, Gabriele-Tergit-Promenade 19, 10963 Berlin, Germany

Attendees (0)