In recent years, product complexity in terms of function and structure has been driven by technological development in complementary components. Designing unbiased product evaluation metrics being to grasp the complex relationships of product features, and able to capitalize on market needs has become a challenge in industrial practice.
In this paper, we propose a hybrid framework in which evaluation models are generated by integrating Interpretive Structural Modeling (ISM), Hierarchical Clustering and Data Envelopment Analysis (DEA). Whereas ISM constructs hierarchical digraphs (skeletons), Hierarchical Clustering reduces dimensionality of pairwise comparisons (correlations) of design variables, and suggests possible evaluation configurations, and DEA computes weights to provide optimal evaluation metrics. Our computational experiments using more than twenty thousand vehicles from 1982 to 2013 confirmed the feasibility and usefulness of DEA with hierarchical concepts to generate the optimal vehicle evaluation metric, and to suggest configurations for vehicle design layouts.