Immersive Analytics touches on so many areas and disciplines. And, so much is happening in those areas every week! Therefore, there are so many IA blogs to be written!
Help! The IA community needs 10-20 regular bloggers who commit to one blog per month. We may be able to track a small fraction of IA happenings.
Here is my list of blogs-to-be-written, which can be expanded into mini-projects suitable for practitioner articles or research papers. Anyone want to co-author any of these? Or, does this list stimulate some blogs of your own? Regardless, let’s start cranking! Email ideas to editors@ImmersiveAnalytics.com.
- Gamifying Data Worlds: Serious games are fascinating in their engagement and education, plus a bit of crowdsourcing, along with their potential for discovery. See this classification of various types of serious games. Probe the following: Fold.it, A Slower Speed of Light, Galaxy Zoo, Guess the Correlation, Zooniverse, Tomnoh, and many more.
- Discover the Model: Create data world with 5-10 dSpaces generated from skLearn random sample generators, all based on the same model. Provide appropriate analytics. Ask questions about aspects of the model, like dimensionality, noise, natural clusters, etc. Could be used as infrastructure for formal experimental design, along with learning concepts, like bias/variance.
- Why Can’t IA be Like Fold.it? Examine what makes Fold.it work, especially the game objective of minimizing bond energy and its implication for legal interactions. What are the equivalents within a data world? Can we apply this interaction style to data worlds?
- Thinking the Unthinkable: There is a wealth of ideas embodied in Bret Victor’s presentation video plus notes, along with his other work. Survey and distill it for data worlds. How do the three mentalities – interactive, visual, symbolic – guide the design of data worlds?
- Trees of Generalizations: Extend the discussion on “data physics” in IEEE IA Workshop paper with a working example of “a learning algorithm would appear has a vigorous forest of trees, constantly striving upward”.
- Is Mean a Valid Analytic: Explore the implications of Anscombe’s quartet for data worlds, such as using calculated probability density functions rather than first/second order summary statistics.
- API Design of scikit-Learn for Machine Learning: Extend the skLearn API Design paper with working examples (in python3 notebook) illustrating structures containing analytic results. Suggest ways of mapping those structures into dObjects within data worlds.
- Glyphs On Demand: I was fascinated by the technique of generating glyphs of metagenomics sequences (shape grammar objects) for the ATLAS in silico project. How could this technique be applied to creating dObjects for data worlds?
- IA with Unity3D – Tools & Tips: Survey the Unity3D ecosystem and highlight tools and techniques valuable for building data worlds. Interview expert Unity developers.
- Milgram Reality-Virtuality Continuum: Survey of recent research articles that cite this work from a collection of citations from Google Scholar alerts.
- Multi-User IA Stacks: Survey of the architectures for supporting interactive multi-user worlds, from metaverses (OpenSimulator, High Fidelity, VirBELLA, AltspaceVR) to gaming engines (Unity networking).
- Procedural Generated Worlds – No Man’s Sky: To be released this June for $60, this game by Hello Games has generated lots of hype. Its unique claim is that all scenes are procedurally generated, implying that you will never see the same twice and that the labor-intensive work of creating static scenes is gone. What are the implications for building data worlds?
- IA Lessons from Second Life: During 2006-2010, a creative professional group (numbering close to a hundred, like SciLands) experimented with 3D data vis at large (virtual scale). It would be useful to interview those pioneers and compile the lessons learned.
- Creating Data Worlds with MineCraft: As a world-building tool, MineCraft has evolved into a community of its own, with many educational use cases. There are several open MineCraft-like gaming engines, like MineTest. Could there tools be used to build simple extensible data worlds?
- IA and Visual Analytics Research: Revisit the goals of the IEEE Visual Analytics community in light of current research and practice. What are the implications for building data worlds?
- Immersive Analogies: Explaining the sense-of-presence of an immersive experience (virtual or otherwise) and its requirements is difficult. And, it is even more difficult to justify why it is useful. Pushing the blog on immersive experiences of cats, what are good analogies for immersive experiences? What are the implications for good designs for data worlds? How can we measure its usefulness?
- Comprehension Power of Data Art: The Art of Analytics hosted by Teradata Corp has a gallery of 26 data art pieces, all generated by analytic algorithms, usually graph networks. Tree of Branch by Todd Margolis is one. All are engaging and beautiful, but what are the implications for designing data worlds to enhance human judgment?
Plus, there are several blog series with specific themes, as suggested below.
- IA Issues: Series on the deeper conceptual issues with building data worlds, but from a practitioner’s perspective. For example, unit of analysis, causality, complex systems, information entropy, etc.
- IA Research: Series on Monash Immersive Analytics Research Group, Caltech Center for Data-Driven Discovery, xREZ Art+Science Lab, and other IA-oriented research groups.
- IA Vendors: Series on Daden, and other vendors of IA-supportive products/services
- IA Inspiration: Series on examples that inspire new ideas about IA, such as Chihuly glass sculptures.
- IA Reflections: Series on reflections of previous work and its implications to IA, such as Turkey’s PRIM-9 interactive analysis, Herbert Simon’s Sciences of the Artificial, and the like.
We will update this list as our ideas mature.