Abstract:

This work presents an approach to generate eLearning materials dynamically from XML-based contents using a XML publishing environment and a text-to-speech system in different contexts and modalities.
Speech output is particularly important for mobile services and uses, because it requires the least attention from the user to technical components, but allows them to focus on the contents of the presentation.
Synthetic speech offers a simple and flexible alternative to recording professional announcers in a studio. Speech output makes sense when used in mobile applications, as it solves the problem of small displays and the difficulty of displaying certain items.
The use of XML as a medium-neutral data format for data retention and processing is structured to be able to produce eLearning content in a dynamic fashion without additional manual efforts, allowing tailoring to different user requirements and optimized to publish in given situations. Contextualization and multimodality of inputs and outputs play a central role for optimal adjustment and use of eLearning in different learning scenarios and learning situations.

Keywords:

multimodal, m learning, dynamic transcoding, text to speech, speech synthesis.