A component framework for personalized multimedia applications

Multimedia content today can be considered as the composition of media elements in time and space into a coherent, interactive multimedia presentation. Personalization of such multimedia content means that it needs to reflect user’s profile information and context information. To create such personalized multimedia content, a manual authoring of many different documents for all the different users’ needs, however, is not feasible not to mention economical. Rather a dynamic authoring process of selecting and assembling personalized multimedia content depending on the user’s profile information and context information seems reasonable. A practical support for such a dynamic authoring of personalized multimedia presentations is neither provided by industrial solutions nor research projects today. To provide a general and at the same time practical support for the dynamic authoring process, a software engineering approach is pursued with the MM4U (short for “MultiMedia For You”) framework. The aim of the MM4U component framework is to provide application developers with an extensive and domain independent support for the authoring of personalized multimedia content. The framework provides generic functionality for typical tasks of the general process chain for creating personalized multimedia content. This includes selecting media elements such as audio, video, image, and text according to the user’s individual profile and context information. These media elements are then assembled and composed in time and space using an abstract multimedia content representation model. This model captures the different aspects of the multimedia presentation such as the temporal course, the spatial layout, and the interaction possibilities without instantiating these in a concrete syntax and format. Only in the subsequent transformation phase—called the last mile—the multimedia content is transformed into a concrete multimedia presentation format, like SMIL, SVG, and Flash, and delivered to the targeted (mobile) end device. To reach the goal of a software framework providing generic support for developing personalized multimedia applications raises the question of a proper software engineering support to develop such a framework. A software framework, like the MM4U framework, typically constitutes a semi-finished software architecture for a specific domain. Since the introduction of object-oriented frameworks in the late eighties, the development of software frameworks is still costly and difficult to handle. To reduce development risk, process models and development methodologies have been developed. With the emergence of component technology also so-called component frameworks appeared. Component frameworks aim to overcome some drawbacks of object-oriented frameworks, such as problems that occur when combining object-oriented frameworks and changes at the bases classes of these frameworks become necessary. In contrast to object-oriented frameworks, a proper process model for component frameworks is still missing. To improve the development process of component frameworks, the dissertation contributes the ProMoCF approach (short for “Process Model for Component Frameworks”), a lightweight process model and development method for component frameworks. This process model has been developed in mutual benefit with the development of the MM4U component framework. While the ProMoCF approach has been evaluated by applying it for the development of the MM4U component framework, the component framework itself has given important feedback for improving and maturing the ProMoCF approach.