Information Visualization
Information visualization is the use of interactive visual representations of abstract data to amplify human understanding, helping people explore, analyze, and communicate information.
Definition
Information visualization is the design of interactive visual representations of abstract data, chosen and arranged to exploit human visual perception so that patterns, relationships, and outliers can be seen and reasoned about.
Scope
This area covers the visual representation of abstract, non-spatial data within human-computer interaction: how data is mapped to visual marks and channels, how perception governs what works, the interaction techniques that make visualizations explorable, the visualization of graphs and networks, and the integration of visualization with analysis in visual analytics. It does not cover scientific visualization of inherently spatial physical data, nor the statistical methods of data analysis themselves, which belong to statistics.
Sub-topics
Core questions
- How is abstract data mapped to visual marks and channels?
- Why do some visual encodings convey information more effectively than others?
- How do interaction techniques support exploration of large datasets?
- How does visualization combine with computation to support analysis?
Key concepts
- visual encoding (marks and channels)
- perceptual effectiveness
- overview, zoom, filter, details on demand
- interaction in visualization
- graph and network visualization
- visual analytics
- data-ink ratio
- exploratory data analysis
Key theories
- Using vision to think
- Information visualization externalizes data into visual form so that the high-bandwidth human visual system can detect patterns and offload cognition, turning perception into a tool for reasoning about abstract information.
- The visual information-seeking mantra
- Shneiderman's principle, overview first, zoom and filter, then details on demand, organizes how interactive visualizations let users navigate from a broad view to specific detail, structured by a task-by-data-type taxonomy.
- Effectiveness of visual encodings
- Visual encoding choices can be ranked by how accurately people read them, and principled design, such as maximizing the data shown relative to ink, produces clearer, more truthful displays.
Clinical relevance
Information visualization helps people make sense of large and complex data in fields from science and finance to public health and journalism; well-designed visualizations support faster, more accurate insight and decision-making, while poor ones can mislead.
History
Building on statistical graphics and cartography, information visualization emerged as a distinct field in the 1990s, consolidated by the 1999 Readings collection and Shneiderman's task taxonomy. Tufte's writing shaped principles of graphical excellence, and later texts such as Munzner's systematized design, while visual analytics arose in the 2000s to couple visualization with automated analysis.
Key figures
- Stuart K. Card
- Jock D. Mackinlay
- Ben Shneiderman
- Tamara Munzner
- Edward R. Tufte
Related topics
Seminal works
- card1999
- shneiderman1996
- tufte2001
Frequently asked questions
- How is information visualization different from scientific visualization?
- Information visualization deals with abstract data that has no inherent spatial form, such as financial records or social networks, so the designer must invent a spatial mapping. Scientific visualization depicts data that is already spatial or physical, such as medical scans or fluid flows, where the geometry is largely given.
- Why does the choice of chart type matter so much?
- Different visual encodings are read with different accuracy by the human visual system; position and length are judged precisely, while area and colour are judged less so. Choosing an encoding matched to the data and task makes patterns easier to see and reduces the risk of misreading.