Visual Representations

Abstract

Visualization can help in the process of extracting insight from data during decision-making. Its advantages are based on the ability to rapidly interpret large quantities of data. The challenge in this context consists in constructing visualization systems which enable the user to understand and perceive the data effectively by providing transparent interaction methods for effective communication between the user and the data. Ware [872] states a number of advantages of visualization, namely:

• The ability to comprehend huge amounts of data.

• The perception of properties that are otherwise not anticipated.

• The extraction of problems with data itself, i.e., detecting outliers or anomalies.

• The understanding of both large-scale as well as small-scale features of data.

• The creation of various hypotheses related to the data.

Visual representations range from simple maps and schematic diagrams to computer generated 2D and 3D visualizations. They differ themselves from other kinds of representations, e.g., sentential representations, in the fact that they make use of visual variables to convey the information to the user. In contrast, sentential representations use the semantics of their constituent units, words, and letters, to achieve the same thing. Thus, the meaning of a word is dependent on its context and its letters but not on the spatial position of the word.

The aim of visual representations is dependent on the context of their usage. However, their advantages are usually focused on these three categories [85]:

• Information recording, that is related to the usage of visual presentations as a storage mechanism in order to avoid the need for memorization of data and their relationships.

• Information communication, where visual representations play the role of a message for communicating the essential features of the data visualized.

• Information processing, where visual representations are used as a means to derive knowledge from data.

Whereas the first two advantages are valid also for other kinds of representations, the last one is specific for visual representations.

This chapter is concerned with visual representations in the context of human-centered visualization environments. Considering the variety of issues that relate to visual representations, it will focus on computer-generated visual representations in different contexts. Before introducing the various techniques which are used to visualize specific data types, a short introduction to perceptual issues is given in Section 4.1. Furthermore, different issues, such as taxonomies for data, visual variables and their ability to express different data types, are discussed in Section 4.1. Section 4.2 focuses on general criteria used in information visualization and the use of metaphors. Section 4.3 presents a survey of different visualization techniques, mainly in the context of multivariate data, by discussing properties of different techniques related to their comprehensive and interaction properties, advantages, and drawbacks with illustrations given for different data sets. Section 4.4 provides an overview of existing evaluations on visualization techniques designed for graphs and trees. The survey mainly focuses on the effect of graph layout algorithms which compute the position of nodes and edges, as well as on user’s understanding of graphs for different tasks. Section 4.5 covers several issues concerning multiple view visualizations. Apart from a classification of multiple view instances, design-issues are discussed. The section is concluded with a comparison based on evaluation studies between multiple views and integrated views.