(This post is a fragment of Chapter 2 of my PhD thesis entitled “Visual Exploration of Web Spaces”)
The Process of Information Visualization by Dürsteler & Engelhardt
The understanding of the basis of data transformation into insight, known as the the process of InfoVis, is crucial for developing effective strategies that help users to reach their informative goals. Several conceptual approximations to such a process have been presented. However, all of them converge in the definition of three main steps, specifically named by Dürsteler and Engelhardt:
• From Data to Information: Once relevant data regarding a target problem has been collected, it has to be structured and organized in order to be transformed into information. Three tasks can be identified concerning such a conversion:
the collection and storage of raw data relevant to the object of study
the processing and transformation of such data to filter errors. This process implies the deletion of irrelevant and redundant records, and the creation of derived magnitudes.
the usage of metadata to build data tables organized according to their meaning
• From Information to Visual Representation: With the data already transformed into information, this step deals with the conversion of information into a perceptual representation, mainly in a visual form. This perceptual representation has to display the information in such a way that the underlying patterns and structures have to be easily identified. A notational schema, which is a particular visual language that maps information into graphics, must be applied to take advantage of previous knowledge or experience of the user. There are two kinds of representations, the arbitrary conventional and the sensory. The former is the one learned over time by the receptor and its main characteristic is that it does not have any perceptive base. The latter is based on symbols and visualisation aspects that use the perceptual capacity of the brain without any previous knowledge or experience. To let the visualisation produce a perceptual impact to the user, the designer has to take into account both sensory and arbitrary conventional representations. Moreover, visual perception, cognitive psychology and even linguistics must be considered in order to provide a pleasant and understandable visual experience.
• From Visual Representation to Understanding: Once the visual representation has been built, it has to be given to the receptor. In order to help gain insight and build knowledge, the visualisation must allow the user to interact with it and empower the analytic discourse. In this regard, Shneiderman introduced the visual information seeking mantra: “overview first, zoom and filter, then details-on-demand”. This mantra suggests the need for providing a general overview of the data,that can be filtered, zoomed and modified at any time so the user can get a deeper insight.
Understanding such a process is crucial to create effective visualizations able to assist in the generation of insight.