Evaluation of Domain Modeling Decisions for Two Identical Domain Specific Languages

Decisions on domain model representations and on domain model scoping are a key task when developing software product lines. However, there is little guidance for t he choice whether a domain model should be represented by feature-oriented domain models or by domain-specific models and whether a domain should be described by a single domain model or by multiple domain models. In this paper we apply a method for assessing domain models on two automation system case studies which use domain specific languages (DSLs). The two case studies cover the same domain (warehouse automation) and were developed independently from one another. We evaluate how well the results of the domain model assessment method are suited for our case studies. For the evaluation we consider the development times of the DSLs, the size of the models, and we calculate DSL and code complexity metrics.