ADML: A framework for representing inhabitants in 3D Virtual Worlds

Abstract

Last years, a new fashion over Internet and computer networks has been established: the 3D virtual worlds. Reports about Future Internet predict virtual worlds will play a major role in the configuration of the future network. They bet on those virtual environments will encompass the ability to become highly adaptive to user preferences (having into account eInclusion aspects), and they will provide a high level of interoperability. However, current 3D virtual worlds are far from cover these issues. This work is focused in providing a preliminary solution about customization and interoperability in the field of avatars representation. They are a key element in the 3D virtual worlds because using 3D avatars is a common way for representing virtual world inhabitants. First of all, nowadays there is not a way for defining all the avatar features that are important for simulating a real person, and for defining their identity. Besides, features that can be specified are usually defined in a closed and proprietary format that does not allow migration among virtual worlds. Thirdly, eInclusion aspects are not taken into account. This work proposes a preliminary framework for specifying a complete avatar identity by means of features that are not dependent on the virtual world context or interaction events. Those are appearance, personality and communication skills. Moreover, for providing a tool that describes the framework, the base for a new XML-based language -ADML- has been created. The language will allow specify avatar identity features taking into account eInclusion aspects and it will allow migration among worlds. Part of this language has been accepted into the new MPEG-V standard proposal.