Prerna Chikersal, PhD

Prerna Chikersal, PhD

Machine Learning Scientist II,
Tripadvisor

Machine Learning

Data Science

Ubiquitous Computing

Natural Language Processing

Human-Computer Interaction

pchikersal[@]gmail.com

Personalized Emoji-first Messaging

Emoji are known to have a strong impact on emotions and positively affect relationships. People in close relationships use emoji in very personalized and secretive ways, thereby forming a shared vocabulary. In this project, we present EmoChat, an emoji-first messaging app for pairs of users in close relationships that leverages the pair's closeness and shared experiences to create a highly personalized "communication zone" with the goal of enhancing communication. In EmoChat, each message is fully translated to emoji. The recipient always sees the emoji first, but can view the text on demand with a tap. Through a field study, we investigated the affordances of personalized emoji-first messaging for close relationships, and we found that emoji-first messaging contributes the following values to text messaging: it adds to humor; it builds secrecy for the exchange of private messages; it promotes creativity, gamification, fun, and play; and it promotes companionship through affection and the sharing of perspectives. To the best of our knowledge, EmoChat is the first ever emoji-first messaging app. Our work created a new paradigm for messaging that enables the communication of lightweight symbols before text and has implications for the future ecosystem of communication apps.







Manucript: Chikersal, P., Bayer, K., Nayar, S. K., Smith, B. A., When Emoji Speak Louder than Words: Personalized Emoji-first Messaging for Enhanced Communication between Partners and Close Friends. In Preparation for TBD.

This was my internship summer project at Snap Research.

Snap Inc. filed the following 3 patents based on my work:

Bayer, K., Chikersal, P., Nayar, S. K., Smith, B. A. Client device processing received emoji-first messages. U.S. Patent 11593548, filed Apr 20, 2021, and issued on Feb 28, 2023.

Bayer, K., Chikersal, P., Nayar, S. K., Smith, B. A. Personalized emoji dictionary. U.S. Patent 11531406, filed Apr 20, 2021, and issued on Dec 20, 2022.

Bayer, K., Chikersal, P., Nayar, S. K., Smith, B. A. Emoji-First Messaging. U.S. Patent 20220337540. Patent Pending.

My Contributions: Led the design and implementation of the emoji-first messaging paradigm. Implemented the prototype as an iOS app written in Swift using SwiftUI and TensorFlow Lite. The app design process was iterative where I tweaked the app based on feedback from pilot participants. Conducted a field study with 22 participants, followed by surveys and user interviews. Leveraged grounded theory to open code interview transcripts, and used thematic analysis to identify themes that identified how the users used EmoChat, the affordances that emoji-first messaging offers, and the pain points and limitations. Presented my findings and discussed design implications for future products at Snap with the HCI team. Led paper writing.