Model Sensemaking Strategies: Exploiting Meta-Model Patterns to Understand Large Models

Abstract

The increasing popularity of model-based and lowcode platforms has raised the need to understand large models -especially in industrial settings. However, current approaches mainly rely on graph-based visual metaphors, which do not scale well with large model sizes. To address this issue, we introduce model sensemaking strategies: purposeful model visualisations based on alternative visual metaphors. We define them as reusable patterns that yield tailored visualisations when applied to meta-models. This paper presents a catalogue of domain-specific and domain-agnostic sensemaking strategies, and a recommender that suggests suitable strategies for a given meta-model. To showcase the framework's applicability, we have implemented some of these strategies in Dandelion, an industrial, low-code graphical language workbench for the cloud. Using this platform, we have evaluated the effectiveness of the strategies to visualise large industrial models by the UGROUND company.