A DEMATEL-Guided Agent-Based Simulation Framework for Trait-Driven Neuroadaptive Interfaces in Smart Hospitality

Heidari, Majid · 2025 · Crossref

DOI: 10.21203/rs.3.rs-8108792/v1

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses the methodological gap in smart hospitality systems, where personalization is typically driven by inferred preferences rather than underlying neuro-cognitive traits. The authors argue that current adaptive interfaces fail to account for latent psychological constructs such as emotional reactivity, cognitive load tolerance, and fairness sensitivity, leading to user disengagement or perceived unfairness. To resolve this, the paper proposes a trait-driven neuroadaptive interface framework that aligns system logic with user traits, grounded in Dual-Process Theory and Affective Computing. The research aims to validate how causal modeling of these traits can improve trust, satisfaction, and decision efficiency in AI-mediated tourism contexts. The methodology employs a dual-stage design integrating the Decision-Making Trial and Evaluation Laboratory (DEMATEL) method with agent-based simulation. First, seven experts provided pairwise ratings to construct a direct influence matrix linking three user traits (emotional reactivity, cognitive load tolerance, fairness sensitivity) to three interface demands (simplicity, ethical alignment, information control). DEMATEL analysis calculated prominence and net causality scores to identify causal drivers. Second, these causal structures informed an agent-based simulation involving 500 cognitively differentiated agents. Each agent possessed randomized trait profiles and interacted with one of three interface types: static, personalized, or fairness-aware. Behavioral outcomes, including trust, satisfaction, and cognitive load (measured via decision latency and reversal frequency), were analyzed to assess trait–interface congruence. The DEMATEL results identified Emotional Reactivity and Ethical Alignment as primary causal drivers, with Emotional Reactivity exhibiting the highest prominence (7.1) and positive net causality (0.33). Conversely, Information Control and Cognitive Load Tolerance functioned as passive receivers. The simulation findings indicated that alignment between agent traits and interface logic significantly improved user experience. Specifically, fairness-aware interfaces enhanced perceived fairness and trust more effectively than static or purely personalized systems. Agents with high emotional reactivity reported greater satisfaction when interacting with adaptive interfaces that adjusted affective tone. Furthermore, high cognitive load reduced trust unless mitigated by personalized or fairness-aware features, confirming that structural alignment with dominant traits reduces cognitive overload and confusion. The study concludes that effective personalization requires moving beyond preference matching to incorporate neuro-cognitive variability and ethical transparency. By embedding fairness and emotional responsiveness into system parameters, developers can create interfaces that are psychologically intelligent and ethically accountable. The proposed framework offers a replicable process for trait-based adaptation, with implications extending beyond hospitality to other AI-mediated domains such as healthcare and education. This approach advances the field by providing a computational and methodological rationale for designing systems that adapt to the diverse cognitive and emotional needs of users.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-nvidia on 2026-08-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• automation