Abstract

Although research on Model-Driven User Interface Develop- ment (MD-UID) has started since the 80’s, with ubiquitous spread of dierent types of devices (e.g. smart phones, PDAs, etc.), this approach has been focused more by researchers in recent years. Nowadays, an ap- plication to be accepted by end users should be compatible with more devices than before. These devices have varieties in dierent aspects such as screen size, IO device and etc. and as a result an application running on them should be compatible with these characteristics. A basic solu- tion to deal with these varieties is developing a speci c application with speci c user interface per device. It is obvious that this approach is full of redundant eorts. The goal of MD-UID is to provide an environment or language to describe the user interface speci cation in the higher level of abstraction to capture the commonalities between various devices. The resulting artifact of this phase can be transformed to a concrete model compatible with the target device afterwards. In this survey, we explain dierent approaches in MD-UID and enumerate the advantages and dis- advantages of them.