Trust Me, I'm Funny: Humor, Personalization, and Trust in Conversational Agents
A Systematic Literature Review on User Engagement, Educational Adoption, and Responsible Use
Brîncoveanu, C. (2025). Trust Me, I’m Funny: Humor, Personalization, and Trust in Conversational Agents – A Systematic Literature Review on User Engagement, Educational Adoption, and Responsible Use. In INFORMATIK 2025. Gesellschaft für Informatik eV.
External URL: https://doi.org/10.18420/inf2025_83⤴
Download: PDF⤴
Abstract: As conversational agents (CAs) permeate society and transform fields like digitized learning, ensuring their responsible and sensible application is crucial. This Systematic Literature Review (SLR) examines how six conversational design dimensions shape user experiences and influence the potential for responsible Artificial Intelligence (AI) use, with a particular lens on educational settings. Using Wolfswinkel et al.’s guidelines to synthesize findings from 88 relevant studies, we analyze how each identified dimension impacts engagement, adoption, and perceived credibility. The results highlight benefits from features like tailored humor and personalization for enhancing engagement, but also reveal significant challenges related to context sensitivity, ethical boundaries, and maintaining trust. These insights inform theoretical models and yield practical recommendations for designing trustworthy, adaptive conversational agents that can be effectively deployed in educational contexts and contribute positively to the societal implications of AI in an open digital society.
Keywords: Conversational Agents, Humor, Anthropomorphism, Personalization, Proactivity, Trust, Ethics, Systematic Literature Review
BibTeX:
@incollection{incollection,
author = "Brîncoveanu, Constantin",
title = "Trust Me, I’m Funny: Humor, Personalization, and Trust in Conversational Agents – A Systematic Literature Review on User Engagement, Educational Adoption, and Responsible Use",
year = 2025,
doi = "10.18420/inf2025_83",
booktitle = "INFORMATIK 2025",
publisher = "Gesellschaft für Informatik e.V.",
address = "Bonn",
pissn = "2944-7682",
pages = "965",
}