Online Mingling: Supporting Ad Hoc, Private Conversations at Virtual Conferences

Title of the Paper

Online Mingling: Supporting Ad Hoc, Private Conversations at Virtual Conferences

Paper Information

• Subject Area: Social interaction and support for spontaneous communication in virtual conferences
• Keywords: Virtual conferences, social interaction, ad hoc conversations, video conferencing, online interaction

Research Background and Issues

1. Problems and Challenges:

   • Current video conferencing systems (e.g., Zoom, Skype) are highly effective for formal meetings but fail to adequately support spontaneous, private interactions such as pre-meeting, post-meeting, and inter-session "hallway conversations."
   • These spontaneous interactions are critical for fostering creativity, innovation, social bonds, and trust.
   • The COVID-19 pandemic has accelerated the adoption of virtual conferences, but the lack of support for casual interactions has led to lower user satisfaction with virtual meetings.

2. Significance:

   • Spontaneous interactions play a crucial role in innovation (e.g., generating new knowledge, breakthrough discoveries) and social connections (e.g., fostering trust and a sense of belonging).
   • The reduction of in-person meetings (to lower carbon emissions and address health risks) makes it increasingly important to improve virtual meetings to support ad hoc conversations.

3. Research Motivation:

   • Surveys indicate that interacting with others is one of the most important aspects of in-person conferences, yet it is poorly supported in virtual settings.
   • There is a desire to address the lack of spontaneous interactions in virtual meetings through technological solutions, thereby enhancing the appeal and feasibility of remote work.

Solution

1. Proposed Approach:

   • Developed a system called “Minglr” to support spontaneous private interactions in virtual conferences. Minglr simplifies the complex process of initiating conversations by matching conversation partners efficiently.

2. Innovations:

   • Provides user-controlled matching processes instead of fully automated ones.
   • Supports random queues and multiple selections (users can wait for multiple conversation partners simultaneously).
   • Allows users to quickly search and select potential conversation partners based on personal profiles.

3. Implementation Steps and Key Technologies:

   • Registration and Profile Creation: Users log in via Google or Facebook and edit their profiles (including interests and avatars).
   • Finding Conversation Partners: Active user profiles are displayed on the system interface, supporting interest-based user searches.
   • Matching and Interaction: Once a match is confirmed, users are directed to a private video meeting room.
   • Technical Implementation: The system is built on the open-source video conferencing platform Jitsi and is available as open-source software (https://github.com/CCI-MIT/minglr).

Research Outcomes

1. Experimental Data:

   • The system was tested during two virtual conferences:
     • ACM Collective Intelligence Conference (CI 2020) and Microsoft New Future of Work Workshop (NFW).
   • Over 35% of participants used Minglr, generating more than 80 conversations with a total duration of 7.5 hours.
   • The average waiting time was only 9 seconds, demonstrating high efficiency.

2. Survey Results:

   • Among users who experienced the system without technical issues, 81% believed that future virtual conferences should include features like Minglr.
   • Users rated the system highly for ease of use and matching efficiency, though some newcomers and introverted users still faced barriers to interaction.

3. Advantages:

   • Simplifies the process of spontaneous interactions, enhancing connections among participants.
   • The system’s search and randomization features help users find like-minded partners, compensating for the limitations of physical environments.
   • The efficient matching process reduces user waiting times, optimizing the overall user experience.

4. Limitations:

   • Insufficient support for newcomers and introverted users.
   • The system currently supports only two-person conversations, lacking additional features such as group discussions or random pairings.

5. Future Directions:

   • Develop specific icebreaker features, such as interest-based matching mechanisms.
   • Introduce newcomer introduction functions to help experienced users connect with new participants.
   • Improve technical compatibility to minimize the impact of technical issues on user experience.