Lecture Recording: Structural and Symbolic Information vs. Flexibility of Presentation

Abstract

Rapid eLearning is an ongoing trend which enables flexible and cost‑effective creation of learning materials. Especially, lecture recording has turned out to be a lightweight method particularly suited for existing lectures and blended learning strategies. In order to not only sequentially playback but offer full fledged navigation, search and inspection of the recorded lecture, chapter marks and search indices have to be embedded. To solve this, two basic approaches for lecture recording tools can be identified — both of them having certain advantages and drawbacks. On the one hand there are systems based on symbolic representation of common slideshow formats like MS PowerPoint. Therefore, they preserve structure and symbol information contained therein, but are lacking flexibility of supported dynamic and interactive formats. On the other hand there are systems based on pixel representation and screen grabbing technologies. While supporting any presentation content, structural and symbolic information cannot be extracted directly and thus has to be post‑processed from the recorded video. This paper discusses a perspective of combining these approaches by widening the slide‑metaphor to a more flexible scene‑based presentation, preserving both the structural and symbolic information. One possible attempt for this is identified by introducing a browser‑based scene concept. Symbolic information can be directly extracted from the XHTML source code and structural information derives from switching through scenes. The browser itself is capable of presenting a wide range of dynamic and interactive formats, thus offering more flexible presentations. For approving the proposed concepts, a prototype called "Virtual Overhead" was developed and evaluated.