Answers are grounded in this paper and cite their sources.
Purpose This study examines how users interpret emotional cues and social presence in artificial intelligence (AI)-mediated academic library reference services, and how these interpretations are shaped by users' interaction orientations and their attributions of service agents as either human librarians or AI systems. Design/methodology/approach Adopting a qualitative, theory-building approach, this study conducted semi-structured interviews with 22 users of an academic library's online reference services, including AI-based chatbots and human-mediated chat services. Data were analyzed using abductive analysis to explore how users construct interaction expectations, evaluate emotional cues and make strategic choices across service agents. Findings The findings show that emotional cues and social presence do not function as universally positive interactional resources. When the service agent is attributed as AI, users tend to prioritize response speed, informational completeness and functional extensibility, while emotional expressions are often deprioritized or perceived as unnecessary. In contrast, emotional cues in interactions with human librarians enhance trust and engagement but may also generate implicit social obligations and emotional pressure. Users' interaction orientations play a central role in shaping these evaluations and service choices. Originality/value This study reconceptualizes social presence as a conditional and user-differentiated construct in library reference services. By integrating interaction orientation and service agent attribution into a unified framework, it advances theoretical understanding of interaction quality in hybrid human–AI reference service environments.
No takes yet. Share an insight, caveat, or question.
Shih-Chuan Chen (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: