Chapter 1: Introduction Flashcards
Why Keep a human in the loop when trying to solve a problem with data?
For well‐defined questions, computational techniques from statistics and machine learning often suffice and are much more effective than manual computations by humans. However, many analysis problems are ill specified People don’t know how to approach the problem. There are many possible questions to ask. When this occurs, it is often helpful to keep a human in the loop as humans have powerful pattern detection capabilities due to the human visual system.
Why have a computer in the loop?
By enlisting computation, you can build tools that allow people to explore or present large datasets that would be completely infeasible to draw by hand, thus opening up the possibility of seeing how datasets change over time. Computer allow the exploitation of human perception when working with large datasets.
Why use an External Representation, like a visualization on a computer screen?
Visualization allows people to offload internal cognition and memory usage to the perceptual systems, using carefully designed images as a form of external representations. The advantages of external memory is that information can be organized by spatial location, offering the possibility of accelerating both search and recognition
Why Depend on vision?
Visualization is based on exploiting the human visual system, rather than other sensory modalities because it is both well characterized and suitable for transmitting information. The visual system provides a very high-bandwidth channel to our brains and a significant part of processing visual information is done in parallel at the preconscious level. Sound is poorly suited for providing overviews of large information spaces compared with vision. We perceive sound as a sequential stream, rather than as a simultaneous experience where what we hear over time is automatically merged together. The other senses can be ruled out because of technical limitations.
Why show data in detail?
Statistical characterization of datasets is powerful, but it has the limitation of losing information through summarization. A single summarization is often an oversimplification that hides the true structure of the dataset, and this is even more true for large and complex datasets.
Why use interactivity?
A single static view can show only one aspect of the data. By implementing interactivity, more queries can be supported and the user will be able to explore the data in more detail. Interaction support the investigation of the data in multiple levels of detail.
Why is the visualization idiom design space huge?
A visualization idiom is a distinct approach to creating and manipulating visual representations. There are many ways to create a visual encoding of data in a single picture. The design possibilities get even bigger when you consider how to manipulate one or more of these pictures with interaction.
Why focus on tasks?
A tool that serves well for one task can be poorly suited for another, for exactly the same dataset. The task of the users is an equally important constraint for a visualization designer as the kind of data that the user has. Reframing the users’ task from domain-specific form into abstract form allows you to consider the similarities and differences between what people need across many real-world usage contexts.
Why focus on Effectiveness?
The emphasis on effectiveness in visualization is a natural outcome of defining visualization’s goal as supporting user tasks. This objective leads to a strong focus on ensuring the correctness, accuracy and truthfulness of the visualization as they play a central role in successful task completion.
Why are most designs ineffective?
In some cases, a possible design is a poor match with the properties of the human perceptual and cognitive systems, In other cases, the design would be comprehensible by a human in some other setting by it’s a bad match with the intended task. An appropriate goal of designing is to satisfy, that is to one of the many possible good solutions rather than one of the even larger number of bad ones.
Why is validation difficult?
The problem of validation for a vis design is difficult because there are so many questions that you could ask when considering whether a vis tool has met your design goals. examples are questions about the effectiveness, insights, engagement, comparison with other vis systems, how to test for the user, etc. There are many aspects to ‘effectiveness’, depending on your goal.
What types of resource limitations are there?
When designing or analyzing a vis system, you must consider at least three different kinds of limitations: computational capacity, human perceptual and cognitive capacity and display capacity.
1. With computational capacity we have to keep in mind scalability on larger datasets, CPU and memory consumption.
2. On the side of human capacity we have to keep in mind that our own memory is fairly limited for things that are not directly visible. Both for short term and long term recall these limitations come into play, leaving us vulnerable for change blindness.
3. Display capacity is a problem designers usually have to deal with; they usually run out of pixels to show all desired information. the amount of information displayed on a screen can be expressed as the information density, compared to the unused space.
The goal of idiom design choices is to find a balance in the amount of information density.
What is meant by change blindness?
Human memory for things that are not directly visible is notoriously limited. These limits come into play not only for long-term recall but also for shorter-term working memory, leaving us vulnerable to change blindness: the phenomenon where even large changes are not noticed if we are attending to something else in our view.
Why analyse?
It can be hard to decide what to do if you’re confronted with a problem as a vis designer. Looking at evidence-based analysis of different vis tools can help you make satisfying design decisions for your visualization problem.
