An Interactive Metamodel Integration Approach (IMIA) for Active and Assisted Living Systems

Emotion and activity recognition can play an important role in supporting people, especially in the AAL environment. Modeling is one of the important factors that can be used to introduce such support. Metamodel conceptualization provides the basis for designing a modeling framework. Metamodel development is an iterative activity that is commonly evolving over time. This development may need to extend, re-use or integrate a metamodel with others to meet specific requirements. On the one hand, manual integration of a metamodel is time-consuming, tedious and requires special knowledge. On the other hand, the automatic process is challenging, due to the nature and complexity of meta-concepts. Therefore, a method to support metamodel integration is highly desired. In this work, we first consider briefly the fundamentals of modeling from the perspective of cognitive psychology. We then propose an interactive approach to dynamically adapt metamodel creation by allowing the modeler to configure and reuse existing metamodels and generate a woven form. Finally, this approach allows the developer to use such a model in the machine learning activities. We evaluated the integration process against two benchmarks and compared the results. The performance of the proposed approach showed that it enables the generation of a consistent and correct weaved model. It was able to reach about 87% accuracy in retrieving meta-elements.