Predefined Classification for Mixed Mode Environments

Mixed Mode Environments (MME) [1] are environments with different dimensions of heterogeneity: heterogeneous devices, heterogeneous software, and heterogeneous communication technologies. In MME we have many different devices and there are continuously new devices developed and introduced. To deal with these different kinds of devices we as humans intuitively classify them into different types. So that when ever we describe a specific device to somebody who does not know it yet, we first tell what type of device it is. For example when we want to describe an i900 we first tell that it is a smartphone. Similar kind of classification we need for machineto-machine and machine-to-human communication. When humans do this classification, not everyone will do it in the exact same manner. Actually it’s even quite unlikely to have two exactly matching classifications defined by two different persons. Still it is usually not a problem for humans to understand each other. For machines it is different: here they have either to use the same underlying classification or the different classifications used by interacting machines have to be mapped to each other. This matching is not always possible and can require a lot of computation power. Therefore it would be reasonable to have a simple underlying classification to which most more complex classifications could be mapped. In Agent-based Middleware for Mixed Mode Environments (A3ME) [1] we pointed out a predefined lightweight ontology, which has to be used for classification of different devices, capabilities and services. Our initial purpose was to find an existing classification that fits our needs. To check whether those are appropriate, we had to define the requirements (section 2). Afterwards we checked some existing ontologies with our requirements (section 3). In section 4 we describe the definition of the A3ME ontology and its usage.