Evaluating the Complexity of Dynamic Human-Machine Interfaces: Proposal and Validation of a Feature Similarity Value Method

作者：Mu Tong, Zhelin Li, Zhe Zhang, Siushing Man · 发表于：International Journal of Human-Computer Interaction · 年份：2026 · DOI：10.1080/10447318.2026.2639825 · 被引用次数：1 · 研究领域：Motor Control and Adaptation、Time Series Analysis and Forecasting、Ergonomics and Musculoskeletal Disorders

In the fields of information interaction design, the scientific assessment of human-machine interface complexity remains a crucial issue. This study proposed a cognitive complexity calculation method based on the Feature Similarity Value (FSV), which integrated the limited cognitive resource and visual saliency theory. The method quantified the feature differences and categorical diversity between targets and distractors within an interface. An experiment involving 20 participants was conducted using a dynamic interface, collecting behavioral, subjective and eye tracking data to evaluate the method. The results showed that interfaces with higher FSV values led to significantly longer TCT, increased AFD, and larger MSA, indicating elevated cognitive load. Correlation results linked the effects of FSV to changes in eye movement features, suggesting FSV influences cognitive behavior by modulating search strategies. Overall, the FSV method provides an effective tool for predicting cognitive complexity in dynamic HMI and supporting early-stage interface design.