Data Visualization and Information Visualization
Visualization is one of the key tools for gaining insights into data. Human beings make sense of their environment and situations primarily (although not exclusively) from visual input. As has been known for centuries, visualizations of data complement and enhance people’s cognitive capabilities. Computer algorithms, programs, and technologies have greatly facilitated the visualization of complex data through the rapid generation of plots and images and the providing of interactive capabilities that are difficult to attain with static displays, such as visualizations on paper. Interactive visualizations are especially beneficial in data exploration and in analyzing and interpreting complex results. Humanities scholarship is increasingly utilizing visualizations for a variety of research questions. They are widely employed in distant reading and are increasingly used and adapted for close reading analysis as well (Jänicke et al., 2015).
High-dimensional data from analysis or from machine learning models can be visually explored in 2D using algorithms such as t-SNE (t-distributed stochastic neighbor embedding). Visualizations and interactive interfaces are important in cultural heritage and cultural analytics to analyze and explore large collections of images and other visual resources. Data visualization comprises multiple categories. Traditional approaches, such as line graphs, bar charts, pie charts, scatter plots, and bubble charts can be used to represent, analyze, and interpret relatively simple numerical data. Scientific visualization employs advanced 2D or 3D techniques to display the complex, often high-dimensional data frequently found in the physical and social science, engineering, and mathematics. In the humanities, however, data are usually qualitative, categorical, or otherwise non-numerical, and therefore other approaches are needed. Information visualization is the name given to visualizing abstract, non-numerical data to enhance human cognition. Text, geographic locations, and relationships between objects are examples of abstract data represented in information visualizations.
In the digital humanities, t ag clouds, or word clouds, are examples representing the relative importance of words in a text. Graphs, especially network visualizations, are commonly used to express social relationships and connections between entities. A well-known interactive network visualization that shows the nexus, relationships and influences of abstract artists is hosted on The Museum of Modern Art website. Other information visualizations include tree diagrams, sunburst plots, and maps. Visualization software is readily available as commercial packages and open-source distributions. Many libraries and packages are available for Python and R, and new, innovative visualization techniques can be developed and adapted in these languages by humanities scholars. Because of its importance in sense-making and acquiring new insights, information visualization in the digital humanities is an active research area (Drucker, 2011), (Jänicke et al., 2015), (Theron & Fontanillo, 2015), (Brüggemann et al., 2020). Refer to Stefan Jänicke’s website for examples of advanced digital humanities visualizations.
With the advent of “Big Data” and the development of methods and database technologies to address the issues inherent with it, another visualization technique, visual analytics, was developed to emphasize the role of analysis for these large data volumes. Visual analytics is “the science of analytical reasoning facilitated by interactive visual interfaces” (Thomas & Cook, 2006). The main factors in visual analytics are the affordance of a high degree of interactivity to enable exploration of large, complex data, and the involvement of human cognitive capabilities for sense-making, discovery, and gaining insights.