Visual Performance Evaluation Model Construction and Generative AI Prompt Template Design for OA Form Interfaces in Smart Office Scenarios
Economic Society and Humanities · 2026-08 · 3 (8) · 10-22
DOI: 10.62381/e264802 ↗Kurz gefasst
Fachartikel in der Zeitschrift Economic Society and Humanities, 2026-08. Verfasst von Ying Liu und 1 weiteren Autor:innen. Die Arbeit nennt 29 Literaturangaben.
Abstract (Originalsprache)
OA form interfaces serve as critical nodes for human-computer interaction in smart office scenarios, and their visual design quality directly affects user operational efficiency and cognitive experience. Addressing the gaps in existing evaluation research—namely the lack of systematic weight derivation and operational scoring criteria tailored to the B-end scenario of OA forms, and the absence of executable visual design standards for generative AI tools—this study integrates literature review, vendor case analysis, questionnaire survey (n=386), and user interviews (n=12) to construct a visual performance evaluation system for OA form interfaces comprising 5 dimensions and 15 indicators. The Analytic Hierarchy Process (AHP) is employed to derive indicator weights: Information Architecture (0.309) > Layout Rationality (0.234) > Visual Clarity (0.215) > Interaction Usability (0.168) > Aesthetic Consistency (0.074). Pairwise comparison weights exhibit significant ranking differences relative to direct scoring, revealing that information architecture and layout constitute the structural priority dimensions of visual performance. Furthermore, graded checklist scoring criteria are developed for the 15 secondary indicators, and a Visual Performance Evaluation (VPE) linear weighted model is constructed (inter-rater reliability ICC = 0.87). The checklist criteria are transformed into 22 prompt parameters through direct mapping, combinatorial mapping, and constraint mapping, forming a standardized AI prompt template. A controlled experiment (n=54) demonstrates that using the template increases the mean VPE score of AI-generated interfaces by 27.8% (p < 0.001), with more pronounced improvements in visual dimensions whose parameters exhibit higher executability. This study provides an operational quantitative evaluation tool for OA system interface design optimization and offers a methodological reference for translating design evaluation knowledge into generative AI prompts.
Zitieren
APA
Liu, Y., & Hu, Z. (2026). Visual Performance Evaluation Model Construction and Generative AI Prompt Template Design for OA Form Interfaces in Smart Office Scenarios. Economic Society and Humanities, 3(8), 10-22. https://doi.org/10.62381/e264802
Vancouver
Liu Y, Hu Z. Visual Performance Evaluation Model Construction and Generative AI Prompt Template Design for OA Form Interfaces in Smart Office Scenarios. Economic Society and Humanities. 2026;3(8):10-22. doi:10.62381/e264802
BibTeX
@article{Liu2026,
author = {Liu, Ying and Hu, Zhigang},
title = {Visual Performance Evaluation Model Construction and Generative AI Prompt Template Design for OA Form Interfaces in Smart Office Scenarios},
journal = {Economic Society and Humanities},
year = {2026},
volume = {3},
number = {8},
pages = {10--22},
doi = {10.62381/e264802}
}
EUVIMED