Nice read and also food for thought, thanks Enrico! Lately i stumbled upon https://huyen-nguyen.github.io/maker/index.html#gallery - a so called WordStream "A Lightweight End-to-end Visualization Platform for Qualitative Time-series Data" Great work and paper by Nguyen in my opinion which bridges the gap and ease the process of extracting insights from temporal patterns in text data.
In the qualitative-data-equals-categorical data sense, "visualizations that express combinations of concepts that are exclusively qualitative" possibly corresponds to Michael Friendly's "Visualizing Categorical Data" ?
"Symbols that carry meanings" can also arguably include simple data symbols and simple visual channels, e.g., a (possibly outdated) correspondence between blue/pink data points and male/female groups or blue/red for cold/hot.
Yes. It may correspond to "Visualizing Categorical Data." I am not too familiar with it.
Yes! I agree that even simple symbols and specific combinations of channels can carry meaning (and emotions!). Color is very powerful. Shape is also very intricate and expressive.
Great post; thanks for sharing! If you're interested, my thesis (esp. Chapter 3) is about visualizing categorical data: https://hdl.handle.net/10289/17049
Thank you for sharing your ideas - it was fascinating to read the directions you are developing.
I feel you are touching deep questions about how a valid data visualization should really be constructed, and how we can understand data in a more essential way.
Reading your post, it occurred to me that there might be a slight mix between the properties of the measured dataset (such as "Who/What?", "Where?", and "When?") and the structural building blocks required to build a valid description that can be visualized.
In the book I wrote about the **Graphication Model**, I developed a structured framework based on five consistent components:
- **Metric** (what is measured — e.g., number of students)
- **Space** (where it is measured — List, Sequence, Total, Timeline, or Map)
- **Time** (when it is measured — either a Point in Time or a Time Period)
- **Format** (how the data is organized — Ranking, Ordering, Separating, Changing, Placing)
- **Meaning** (what kind of relationship is revealed — Hierarchy, Seriality, Dominance, Trend, Density)
If you’d like, I would be happy to send you the model summary again.
It might provide a clear and solid foundation for some of the important questions you are raising — and I would love to hear your thoughts on it.
I'm happy to continue the discussion if you're interested.
Nice read and also food for thought, thanks Enrico! Lately i stumbled upon https://huyen-nguyen.github.io/maker/index.html#gallery - a so called WordStream "A Lightweight End-to-end Visualization Platform for Qualitative Time-series Data" Great work and paper by Nguyen in my opinion which bridges the gap and ease the process of extracting insights from temporal patterns in text data.
This is very useful! Thanks so much for sharing it!!!
In the qualitative-data-equals-categorical data sense, "visualizations that express combinations of concepts that are exclusively qualitative" possibly corresponds to Michael Friendly's "Visualizing Categorical Data" ?
"Symbols that carry meanings" can also arguably include simple data symbols and simple visual channels, e.g., a (possibly outdated) correspondence between blue/pink data points and male/female groups or blue/red for cold/hot.
Yes. It may correspond to "Visualizing Categorical Data." I am not too familiar with it.
Yes! I agree that even simple symbols and specific combinations of channels can carry meaning (and emotions!). Color is very powerful. Shape is also very intricate and expressive.
Thanks for your comments Paul!
Great post; thanks for sharing! If you're interested, my thesis (esp. Chapter 3) is about visualizing categorical data: https://hdl.handle.net/10289/17049
Thank you for sharing your ideas - it was fascinating to read the directions you are developing.
I feel you are touching deep questions about how a valid data visualization should really be constructed, and how we can understand data in a more essential way.
Reading your post, it occurred to me that there might be a slight mix between the properties of the measured dataset (such as "Who/What?", "Where?", and "When?") and the structural building blocks required to build a valid description that can be visualized.
In the book I wrote about the **Graphication Model**, I developed a structured framework based on five consistent components:
- **Metric** (what is measured — e.g., number of students)
- **Space** (where it is measured — List, Sequence, Total, Timeline, or Map)
- **Time** (when it is measured — either a Point in Time or a Time Period)
- **Format** (how the data is organized — Ranking, Ordering, Separating, Changing, Placing)
- **Meaning** (what kind of relationship is revealed — Hierarchy, Seriality, Dominance, Trend, Density)
If you’d like, I would be happy to send you the model summary again.
It might provide a clear and solid foundation for some of the important questions you are raising — and I would love to hear your thoughts on it.
I'm happy to continue the discussion if you're interested.
Yes please share it! It sounds really useful!