PuppetChat: Fostering Intimate Communication through Bidirectional Actions and Micronarratives

Paper Title

PuppetChat: Fostering Intimate Communication through Bidirectional Actions and Micronarratives

Publication Info

• Topic area: Enhancing intimate communication in instant messaging through expressive and reciprocal interactions.
• Keywords: Instant messaging, intimate communication, social presence, micronarratives, reciprocal actions, dyadic chat, personalization, co-performance, shared memory, HCI.

Background and Problem

• Problem / challenge: Current instant messaging tools dilute intimate communication by relying on emojis and reactions that are often vague, one-directional, and lack continuity or personalization. These tools fail to support nuanced emotional expression, reciprocal interaction, and the accumulation of shared memories.
• Significance: Intimate communication is crucial for maintaining relational closeness, emotional support, and shared meaning, especially in long-distance relationships. Addressing these gaps can enhance the quality of mediated relationships.
• Motivation and related work: Prior research highlights strategies like awareness, expressivity, physicalness, joint action, and memories to improve mediated intimacy. However, existing systems often rely on one-way indicators, lack turn-taking mechanisms, and fail to integrate shared narratives or physical metaphors of care. This paper builds on these insights to create a system that fosters reciprocal and expressive communication.

Solution

• Proposed approach: PuppetChat, a dyadic messaging prototype that combines embodied puppet actions with personalized micronarratives to enhance reciprocal and expressive communication.
• Novelty:
  1. Integration of embodied actions with AI-generated, editable micronarratives for personalized and expressive communication.
  2. A reciprocity-aware recommendation system that suggests complementary actions to encourage co-performance.
  3. Mechanisms for preserving and revisiting interactions to support shared memory and continuity.
• Procedure and key techniques:
  • PuppetChat uses a library of 42 expressive puppet actions paired with short, customizable micronarratives.
  • A scoring-based recommendation system suggests context-aware reciprocal actions.
  • Users can edit AI-generated micronarratives to personalize their messages.
  • Interactions can be ephemeral or preserved as revisitable traces, supporting reflection and shared history.

Results

• Concrete findings:
  • Participants sent an average of 14.5 actions, with 57% accompanied by micronarratives.
  • High usability (SUS score: 86.59) and low workload (NASA-TLX score: 31.72).
  • Positive ratings on co-presence (5.3/7), reciprocal interaction (5.1/7), and agency (5.4/7).
• Advantage over baselines:
  • PuppetChat transformed isolated reactions into reciprocal, co-performative sequences.
  • Editable micronarratives enabled nuanced self-expression and the creation of shared symbolic codes.
  • Persistent interaction records supported continuity and reflection, unlike ephemeral-only systems.
• Experiments / evaluation:
  • A 10-day field study with 11 dyads (22 participants) in close relationships.
  • Mixed-methods evaluation combining quantitative metrics (e.g., action frequency, usability scores) and qualitative feedback (e.g., interviews on social presence and personalization).
• Limitations and future work:
  • Small sample size (22 participants) and short deployment period (10 days).
  • Limited diversity in relationship types and cultural contexts.
  • Future work includes longitudinal studies, cross-cultural deployments, and ablation studies to isolate the impact of specific features.

Summary

PuppetChat enhances intimate communication by integrating embodied puppet actions with editable, AI-generated micronarratives, fostering reciprocal exchanges and shared memories. A 10-day field study demonstrated that the system supports nuanced self-expression, co-performance, and continuity in dyadic messaging. While the study confirmed the system's usability and effectiveness, future research will explore its long-term impact, broader applicability, and specific feature contributions.