Filtered by themeConversational agents8 papersclear ✕

Article Effects of AI Virtual Anchors on Brand Image and Loyalty: Insights from Perceived Value Theory and SEM-ANN AnalysisZhu · 2025AI virtual anchors are an emerging innovation that are gaining significant at- tention, as they hold promising applications across various fields. This study examines how users perceive live product selling by AI virtual anchors and its impact on brand image and brand loyalty. A two-stage PLS-SEM and ANN approach was employed to analyze data from a sample of 336 individuals in China who had experienced and utilized AI virtual anchors for purchases during branded live streaming sessions. The findings indicate that perceived usefulness, perceived enjoyment, and novelty positively impact brand image, with artificial neural network (ANN) analysis identifying brand image as the primary predictor. Furthermore, brand image acts as a mediator between these user perceptions and brand loyalty. These insights offer brand managers a strategic approach to utilize AI virtual anchors for fostering a positive brand image and building loyal customer bases. The study also contributes to the academic understanding of consumer behavior and brand management in the context of AI.

UX of AI & PersonalizationResearch Papers

The Illusion of Empathy? Notes on Displays of Emotion in Human-Computer InteractionAndrea Cuadr · 2024From ELIZA to Alexa, Conversational Agents (CAs) have been delib­ erately designed to elicit or project empathy. Although empathy can help technology better serve human needs, it can also be deceptive and potentially exploitative. In this work, we characterize empathy in interactions with CAs, highlighting the importance of distin­ guishing evocations of empathy between two humans from ones between a human and a CA. To this end, we systematically prompt CAs backed by large language models (LLMs) to display empathy while conversing with, or about, 65 distinct human identities, and also compare how diferent LLMs display or model empathy. We fnd that CAs make value judgments about certain identities, and can be encouraging of identities related to harmful ideologies (e.g., Nazism and xenophobia). Moreover, a computational approach to understanding empathy reveals that despite their ability to display empathy, CAs do poorly when interpreting and exploring a user’s experience, contrasting with their human counterparts. CCS CONCEPTS • Human-centered computing → HCI theory, concepts and models; Empirical studies in HCI.

UX of AI & PersonalizationResearch Papers

Roles of artificial intelligence experience, information redundancy, and familiarity in shaping active learning: Insights from intelligent personal assistantsWang, Sun · 2023Artificial Intelligence (AI) is increasingly being integrated into educational settings, with Intelligent Personal Assistants (IPAs) playing a significant role. However, the psychological impact of these AI assistants on fostering active learning behaviors needs to be better understood. This research study addresses this gap by proposing a theoretical model to outline and predict active learning dynamics. Data was col- lected from 237 validated questionnaires and analyzed using partial least squares structural equation modeling. Our results confirm most hypotheses advanced in our model, and information redundancy has an unexpected negative and indirect influ- ence on active learning, while perceived familiarity and system quality are positive drivers. Crucial mediators such as perceived usefulness, ease of use, and conveni- ence significantly positively influence active learning outcomes. Interestingly, the relationship between perceived ease of use, perceived convenience, and active learn- ing is positively moderated by AI experience. The most striking and unexpected finding of this study is the preference of university students for familiar systems over high-tech learning methods. This result challenges the common belief that the younger generation is always eager to adopt the latest technology. Instead, our find- ings suggest that students value convenience and familiarity over novelty in learn- ing systems. This preference is reflected in their systematic evaluation, where con- venience and familiarity are considered top priorities. This study provides valuable insights into the potential of AI to enrich the learning experience, thus making it especially relevant to professionals interested in artificial intelligence in interna- tional business education.

UX of AI & PersonalizationResearch Papers

Collaboration with Conversational AI Assistants for UX EvaluationEmily Kuang · 2023EMILY KUANG, Golisano College of Computing and Information Sciences Rochester Institute of Technology, USA EHSAN JAHANGIRZADEH SOURE, School of Computer Science University of Waterloo, Canada MINGMING FAN∗ , Computational Media and Arts Thrust The Hong Kong University of Science and Technology (Guangzhou), China and Division of Integrative Systems and Design & Department of Computer Science and Engineering The Hong Kong University of Science and Technology, China

UX of AI & PersonalizationResearch Papers

Zero-shot multitask intent and emotion prediction from multimodal data: A benchmark studyMauajama Firdaus b, Dushyant Singh Chauhan c · 2023Empathy involves comprehending and sharing the emotions of another person. In the realm of conversational AI, empathy pertains to the AI’s capacity to understand and respond suitably to the user’s emotions and needs. Conversational AI with empathetic capabilities can heighten the user experience by making interactions more personalized and natural. At present, machine learning algorithms are commonly utilized in existing conversational AI systems to recognize emotions and corresponding empathetic intents from annotated data. Nonetheless, this approach is not without limitations, being expensive and time-consuming. Our present work takes a holistic approach to empathy in conversational AI, where we propose a novel zero-shot multitask framework, the Zero-shot Intent Emotion Detection (ZIED) network, identifies both emotions and intents in a multimodal setting. We developed an end-to-end model that concurrently captures textual, audio, and visual representations and integrates the different modalities using cross-attention mechanisms. Our experimental results, based on the EmoInt-MD dataset, show that incorporating all three modalities results in the best performance for both emotion and empathetic intent detection. We observed a noteworthy improvement of over 6% and 4% for intent and emotion, respectively, for various ratios of seen and unseen classes.

UX of AI & PersonalizationResearch Papers

User acceptance on content optimization algorithms: predicting filter bubbles in conversational AI servicesLONG PAPER · 2022As the role of artificial intelligence (AI) agents in information curation has emerged with recent advancements in AI tech- nologies, the present study explored which users would potentially be susceptible to the filter bubble phenomenon. First, a large-scale analysis of conversational agent users in South Korea (N = 2808) was conducted to investigate the relative importance of content optimization algorithms in shaping positive user experience. Five user clusters were identified based on their information technology proficiency and demographics, and a multiple-group path analysis was performed to com- pare the influences of content optimization algorithms across the user groups. The results indicated that the personalization algorithm generally exhibited a stronger impact on evaluations of an AI agent’s usefulness than the diversity algorithm. In addition, increased user age and greater Internet usage were found to decrease the importance of objectivity in shaping trust in AI agents. This study improves the understanding of the social influence of AI technology and suggests the necessity of segmented approaches in the development of AI technology.

UX of AI & PersonalizationResearch Papers

Human-Machine Interaction Personalization: aMonica La Mura, Patrizia Lamberti · 2020The increasing spread of pervasive technology has led to the fast development of human-centered connected systems, such as cloud-based voice services, assisted driving systems, domotics control systems, personal digital assistants. The user interacts with these systems by speaking to an artificial intelligence, which interprets the speaker’s requests and takes decision accordingly. In such scenario, the real-time collection of personal information from the speaker’s voice is a key- function to develop in order to offer personalized services. Gender is part of the basic information needed to customize the user experience. Furthermore, knowledge about the sex of the speaker also proves useful in automatic speaker recognition and voice-based identity recognition systems, since it restricts the search space to individuals of one gender, thus speeding up the system response. Therefore, gender recognition techniques through speech analysis have largely attracted the researchers’ attention. Speech analysis is usually performed by extracting some features from the speech signal that can be affected by additional factors other than the gender: emotional state of the speaker, for example, is conveyed in the speech by altering some parameters that take part to the gender recognition process. At the same time, the outcome of emotion recognition systems based on speech analysis can be affected by the speaker’s gender. This paper briefly summarizes the techniques used to perform gender recognition through speech analysis and proposes a practice to take gender into account in emotion recognition methods.

UX of AI & PersonalizationResearch Papers

UX Research on Conversational Human AI interaction - A Literature Review of the ACM Digital LibraryQingxiao Zheng · 2013Early conversational agents (CAs) focused on dyadic human-AI interaction between humans and the CAs, followed by the increasing popularity of polyadic human-AI interaction, in which CAs are designed to mediate human-human interactions. CAs for polyadic interactions are unique because they encompass hybrid social interactions, i.e., human-CA, human-to-human, and human-to- group behaviors. However, research on polyadic CAs is scattered across different fields, making it challenging to identify, compare, and accumulate existing knowledge. To promote the future design of CA systems, we conducted a literature review of ACM publications and identified a set of works that conducted UX (user experience) research. We qualitatively synthesized the effects of polyadic CAs into four aspects of human-human interactions, i.e., communication, engagement, connection, and relationship maintenance. Through a mixed-method analysis of the selected polyadic and dyadic CA studies, we developed a suite of evaluation measurements on the effects. Our findings show that designing with social boundaries, such as privacy, disclosure, and identification, is crucial for ethical polyadic CAs. Future research should also advance usability testing methods and trust-building guidelines for conversational AI.

UX of AI & PersonalizationResearch Papers