AniBalloons: Animated chat balloons as affective augmentation for social messaging and chatbot interaction

Pengcheng An, Chaoyu Zhang, Haichen Gao, Ziqi Zhou, Yage Xiao, Jian Zhao

TL;DR

This paper introduces "AniBalloons," a novel set of 30 animated chat balloon designs intended to convey emotions in text-based communication. Through a series of studies, the research demonstrates that these animations are effectively recognized, can enhance perceived emotional communication quality, and positively impact user experience in both instant messaging and chatbot interactions.

What Problem Does This Paper Address?

Despite the ubiquity of text-based communication, it inherently lacks the rich nonverbal cues present in face-to-face interactions, leading to limitations in conveying emotions and fostering interpersonal connection. While emoticons and stickers offer some augmentation, they are often static or limited in scope. This paper addresses the gap in exploring how animations within the chat balloon itself can serve as a dynamic and expressive channel for affective communication, aiming to enrich the user experience beyond static text and existing emoji solutions.

Methods

The research employed a multi-study, design-driven approach:

  • Study 1 (Design & Requirements): A structured, iterative design process was used to create AniBalloons. This involved defining design requirements (e.g., compatibility with interfaces, complementarity with emojis, non-intrusiveness, readability), collecting and refining design inspirations from existing animation examples, and iteratively designing and testing animations with expert feedback.
  • Study 2 (Affect Recognizability & Properties): 40 participants evaluated the 30 designed AniBalloons for emotion recognizability (how well the intended emotion was perceived without context) and their perceived emotional properties (valence and arousal).
  • Study 3a (Mobile Messaging Experience): 72 participants engaged in a pre-scripted conversation using either a standard messaging interface or one augmented with AniBalloons to assess the impact on perceived affective communication quality, nonverbal information conveyance, message comprehensibility, and interpersonal closeness.
  • Study 3b (Chatbot Experience): 70 participants interacted with a simulated customer service chatbot, again comparing a standard interface with one using AniBalloons, to evaluate the impact on chatbot attitude (fun, likability), emotional expressivity, verbal communication ability, and perceived personality.

Datasets

  • Design Inspiration: 336 motion graphic designs (GIF/MOV) were collected from platforms like Dribbble, Behance, and Pinterest.
  • Evaluation Data:
    • Study 2: Data from 40 participants evaluating 30 AniBalloons.
  • Study 3a: Data from 72 participants in a mobile messaging context.
  • Study 3b: Data from 70 participants in a chatbot context.
  • Reference Data:
    • Frequently Used Emojis (LEED): Used to contextualize AniBalloons' emotional space.
  • VibEmoji Motions: A prior set of animated emoticon designs used for comparison.

Key Insights and Ideas

  • Animation as a Dynamic Affective Channel: The core idea is that animating the chat balloon itself, rather than just its content, can provide a richer, more dynamic way to convey emotions, complementing static cues like emojis.
  • Complementarity, Not Replacement: AniBalloons are designed to work with existing communication methods (emojis, text), not to replace them, aiming to enhance the overall expressive capacity.
  • Abstract Design for Universality: By focusing on motion patterns and abstract effects rather than specific facial features, AniBalloons aim for broader applicability and reduced potential for stereotyping or cultural misinterpretation.
  • Impact on Chatbot Personality: Beyond just conveying emotions, animated balloons can influence how users perceive a chatbot's personality, particularly traits like extraversion.

Key Results

  • Emotion Recognizability: 80% (24 out of 30) of AniBalloons achieved >50% recognition accuracy, with 20 designs reaching >65% accuracy, indicating effective emotion conveyance.
  • Mobile Messaging: Animated chat balloons significantly improved perceived affective communication quality (richness, liveliness, expressiveness), nonverbal information conveyance, and interpersonal closeness. They did not significantly impact message comprehensibility.
  • Chatbot Interaction: AniBalloons made chatbots more fun and likable, increased perceived emotional expressivity, and influenced personality traits (especially extraversion). Again, verbal communication ability was unaffected.
  • Valence-Arousal Distribution: The designed animations covered a broad range of the valence-arousal plane, offering nuanced options for expressing emotions.

Limitations and Future Work

  • Mockup Interfaces: Studies were conducted using high-fidelity mockups, not fully implemented systems. Future work should involve deployment in real-world, longitudinal settings.
  • Selection Interface: The study did not deeply explore the user experience of browsing and selecting animations within a keyboard interface.
  • Accessibility: The accessibility of these animations for users with disabilities was not addressed.
  • Topic Diversity: The conversational scenarios used were limited. Broader examination across various topics is needed.
  • Nuanced Emotion Expression: While basic emotions were covered, future work could explore more complex or nuanced emotional states.
  • User Control: The study explored both user-selected and automatically applied animations. Further research could investigate optimal levels of user control and recommendation mechanisms.

Summary updated

BibTeX
@article{an2025aniballoons,
  author = {Pengcheng An and Chaoyu Zhang and Haichen Gao and Ziqi Zhou and Yage Xiao and Jian Zhao},
  title = {{AniBalloons: Animated chat balloons as affective augmentation for social messaging and chatbot interaction}},
  journal = {Int. J. Hum. Comput. Stud.},
  year = {2025},
  doi = {10.1016/j.ijhcs.2024.103365},
}
Figure