Jan 2008 Multimedia, Graphics and Visualization - Next Steps

Algorithm animation of a 3D force directed layout of DWA512 (Matrix Market) built as an interactive VRML animation.

I intend this new course to focus on Information Visualisation as the driver for the exploration of issues in Multimedia, Graphics and Visualization. I aim to equip the students with a solid grounding in mathematical and algorithmic details while ensuring they can rapidly deliver applied visual tools for exploring voluminous data sets (using the Processing framework which I will cover).

The old course drew on the following texts. Computer Graphics, Principles and Practice by Foley, van Dam, Feiner, Hughes. Principles of Three-Dimensional Computer Animation by Michael O'Rourke the VRML 2.0 Sourcebook by Andrea L. Amea, David R. Nadeau and John Moreland. Graph Drawing: Algorithms for the Visualization of Graphs by Giuseppe Di Battista, Peter Eades, Roberto Tamassia, Ioannis G. Tollis, Information Visualization : Perception for Design by Colin Ware and The Visual Display of Quantitative Information; Visual Explanations: Images and Quantities, Evidence and Narrative; Envisioning Information by Edward R. Tufte.

3D rectangular tree-map of a hard disc. Hierarchy reresented by inclusion, each rectangular box is a directory, size represents voume of data and height represents distance from the root of the file system.

Circular 3D tree-map of a hard disc.

This update to the course takes an innovative approach to the teaching of Information Visualisation in terms of Multimedia, Graphics and Visualization principles. In the teaching and applied learning in this course we will adopt a version of the SECI model(Nonaka & Takeuchi 1995) for knowledge creation. While we will not labour under the SECI model it does provide a framework for identifying a series of check points and guides to the delivery, discussion, active learning and knowledge formation required by students in this course.

In practice we will weekly engage in a seven times through a two week cycle consisting or a series of activities in learning about core concepts in Information Visualisation. Information Visualisation is typically driven by the need to identify questions. We will achieve this through an open-forum with questions and answers, brainstorming, peer work or team work (Socialisation). This forms the tacit to tacit SECI step. Next the course lecturer will engage in a period of tacit to explicit knowledge transfer, helping to convert the tacit ideas discussed earlier into explicit ideas and concepts (including code, mathematical and algorithmic details, infovis methods etc.). This forms the Externalisation step in the SECI model. The next two steps require active student learning and require more time for reflection on the work in the Socialisation and Externalisation steps. A dedicated time for the Combination of the knowledge presented with a specific end-user task is needed. Weekly practical sessions guided by a domain expert or other will aid the students in combining knowledge. Finally, by combining knowledge in this way with a clear task in mind students will start the process of Internalisation whereby explicit knoweldge becomes tacit.

A visual timeline of Winston Churchill’s life, divided into years, months, weeks and days presented on "Lifeline", a fifteen metre-long interactive table in the Churchill Museum London.

Population movement visualisation from "From Migrations to Population Concentration", Gaudin B., Bennett M., Sheehan B. & Quigley A., Best Poster IBM Dublin CASCON 2006

Be attempting to complete this SECI process a number of times (spiral) during this module we aim to move the students up through a series of levels whereby evermore advanced information visualisation concepts are conveyed and realised through practical work. The students will explore the problem inherent in trying to visually display and explore voluminous data sets from sources including web navigation, books, papers/citations, game scores, scientific data, biological data, shopping data, social networks, stock/finance data and news sources. Along with considering the pipeline model of information visualisation we will explore more iterative models dealing with data capture, modeling, filtering, management, processing, refinement, representation and interaction. Students will learn from both research papers and in-class lecture material covering multi-dimensional data, geographical data, biological data, time series data, relational data and methods such as scatter plots, graph drawings and tree layouts.

Labels: infovis, postgraduate