Engineering Adaptive Model-Driven User Interfaces

Software applications that are very large-scale, can encompass hundreds of complex user interfaces (UIs). Such applications are commonly sold as feature-bloated off-the-shelf products to be used by people with variable needs in the required features and layout preferences. Although many UI adaptation approaches were proposed, several gaps and limitations including: extensibility and integration in legacy systems, still need to be addressed in the state-of-the-art adaptive UI development systems. This paper presents Role-Based UI Simplification (RBUIS) as a mechanism for increasing usability through adaptive behavior by providing end-users with a minimal feature-set and an optimal layout, based on the context-of-use. RBUIS uses an interpreted runtime model-driven approach based on the Cedar Architecture, and is supported by the integrated development environment (IDE), Cedar Studio. RBUIS was evaluated by integrating it into OFBiz, an open-source ERP system. The integration method was assessed and measured by establishing and applying technical metrics. Afterwards, a usability study was carried out to evaluate whether UIs simplified with RBUIS show an improvement over their initial counterparts. This study leveraged questionnaires, checking task completion times and output quality, and eye-tracking. The results showed that UIs simplified with RBUIS significantly improve end-user efficiency, effectiveness, and perceived usability.