Group meetings are ubiquitous, and millions of meetings are held every day across the world. However, more than 40% of employees overall do not rate their meetings as productive[1]. Considering the time and money that organizations spend on meetings, improving their quality and effectiveness is crucial. Previous studies have shown that meeting facilitators can be advantageous in helping groups reach their goals more effectively. Given recent advances in AI and Conversational Agent (CA) technology and the effectiveness of CAs in one-on-one interactions, I propose to bring CAs into group settings to act as Virtual Facilitators, and explore ways in which they can facilitate a meeting and provide assistance for a number of tasks in a collaborative, co-located and synchronous multiparty context, such as group decision making.

​

Embodied conversational agents—either virtual or robotic—have several capabilities that can make them effective meeting facilitators, including: the ability to interact with human collaborators in natural language and nonverbal conversational modalities, the ability to hold a non-judgmental and neutral point-of-view, and access to information resources. CAs for short and long-term interactions with individual users have been successfully developed and evaluated. Extending the state-of-the-art to provide group-meeting facilitation requires research to investigate how CAs can improve both users’ subjective perception of meeting outcomes and objective performance. In this proposal, I first review previous research on small-group dynamics and existing challenges of teamwork in workplaces. I then take a closer look at three of the most common challenges affecting group decision-making processes in meetings: a) lack of meeting structure, b) intragroup conflict, and c) implicit bias. After exploring the impact of these challenges on group performance, I present a framework for a CA-based automated group support system to facilitate and structure group decision-making processes, mediate intragroup conflicts, and reduce implicit bias, using a hiring meeting as a working example task. The ultimate goal of this system is to improve the overall quality of the decision-making process and outcome.