First I got to say thank you for taking the time to write about this topic. I find it so captivating and inspiring (I know I mentioned that in your last post comment, but just thinking about the possibilities of how to solve the problems you have been outlining, is like a kid at a candy shop). Second, this maybe rudimentary but I recently stumbled on Neil Theise book "Notes on Complexity" (https://www.perplexity.ai/search/notes-on-complexity-u_o_KpYHT9ay4QKCJClKAA) and I wonder if you think of the connection of scale i.e. how the roads and highways look like blood vessels, how could that approach solve the unbounded space let alone a high number of feature?
Thank you for reading and commenting Omar! I am happy to hear you like the series. I am not familiar with the book but I am fascinated by complexity theory.
Long-time listener, first time caller. I really liked this series and thought I might share a paper with you:
Morison, A.M.; Woods, D.D. Opening up the Black Box of Sensor Processing Algorithms through New Visualizations. Informatics 2016, 3, 16. https://doi.org/10.3390/informatics3030016
This is effectively a "visualization design study" for algorithm output in hyperspectral image analyses, but is grounded in the domain Cognitive Systems Engineering (CSE)... which I find has a lot to offer for visualization research, but sadly seldom makes it way in the discussion. CSE provides a ton of theory that often feels lacking in InfoVis research and speaks to a lot of the issues you have highlighted in other places (e.g., the inadequacy of perceptual-rankings of visual features in describing and informing vis design... or the conceptual mapping between a visualization and the world it represents). I like this particular paper because it nicely weaves a lot of this theory together to an applied problem of opening the "black box" of an algorithm.
Some topics of note:
- Framing the relationship between humans/teams of humans and machines as "joint cognitive systems"
- Describing the architecture of these relationships
- Describing tasks like "corroboration" of algorithm output by depicting relevant "frames of reference" to address the "context-sensitivity problem" (connecting data to the world)
First I got to say thank you for taking the time to write about this topic. I find it so captivating and inspiring (I know I mentioned that in your last post comment, but just thinking about the possibilities of how to solve the problems you have been outlining, is like a kid at a candy shop). Second, this maybe rudimentary but I recently stumbled on Neil Theise book "Notes on Complexity" (https://www.perplexity.ai/search/notes-on-complexity-u_o_KpYHT9ay4QKCJClKAA) and I wonder if you think of the connection of scale i.e. how the roads and highways look like blood vessels, how could that approach solve the unbounded space let alone a high number of feature?
Thank you for reading and commenting Omar! I am happy to hear you like the series. I am not familiar with the book but I am fascinated by complexity theory.
Hi Enrico,
Long-time listener, first time caller. I really liked this series and thought I might share a paper with you:
Morison, A.M.; Woods, D.D. Opening up the Black Box of Sensor Processing Algorithms through New Visualizations. Informatics 2016, 3, 16. https://doi.org/10.3390/informatics3030016
This is effectively a "visualization design study" for algorithm output in hyperspectral image analyses, but is grounded in the domain Cognitive Systems Engineering (CSE)... which I find has a lot to offer for visualization research, but sadly seldom makes it way in the discussion. CSE provides a ton of theory that often feels lacking in InfoVis research and speaks to a lot of the issues you have highlighted in other places (e.g., the inadequacy of perceptual-rankings of visual features in describing and informing vis design... or the conceptual mapping between a visualization and the world it represents). I like this particular paper because it nicely weaves a lot of this theory together to an applied problem of opening the "black box" of an algorithm.
Some topics of note:
- Framing the relationship between humans/teams of humans and machines as "joint cognitive systems"
- Describing the architecture of these relationships
- Describing tasks like "corroboration" of algorithm output by depicting relevant "frames of reference" to address the "context-sensitivity problem" (connecting data to the world)