An Approach for the Automatic Adaptation of Domain-Specific Modeling Languages for Model-Driven Mobile Application Development

Abstract

The use of domain-specific modeling languages (DSMLs) is a common approach to support cross-platform development of mobile applications. However, most DSML-based approaches suffer from a number of limitations such as poor performance. Furthermore, DSMLs that are written ab initio are not able to access the entire range of capabilities supported by the native mobile platforms. This paper presents a novel approach of using an adaptive domain-specific modeling language (ADSML) to support the model-driven development of cross-platform mobile applications emphasizing the Android and iOS platforms. We will discuss the techniques in the design of an ADSML including meta-model extraction, meta-model elevation, and meta-model alignment. We discuss how these techniques can be incorporated into an automated process where a common, platform-independent DSML is dynamically synthesized from the native APIs of multiple target mobile platforms. Our approach is capable of generating high performance native applications; is able to access the full capabilities of the target native platforms; and is adaptable to the rapid evolutions of those platforms.