Visual Insights

Visual Insights: A Practical Guide to Making Sense of Data

Copyright Date: 2014
Published by: MIT Press
Pages: 312
  • Cite this Item
  • Book Info
    Visual Insights
    Book Description:

    In the age of Big Data, the tools of information visualization offer us a macroscope to help us make sense of the avalanche of data available on every subject. This book offers a gentle introduction to the design of insightful information visualizations. It is the only book on the subject that teaches nonprogrammers how to use open code and open data to design insightful visualizations. Readers will learn to apply advanced data mining and visualization techniques to make sense of temporal, geospatial, topical, and network data. The book, developed for use in an information visualization MOOC, covers data analysis algorithms that enable extraction of patterns and trends in data, with chapters devoted to "when" (temporal data), "where" (geospatial data), "what" (topical data), and "with whom" (networks and trees); and to systems that drive research and development. Examples of projects undertaken for clients include an interactive visualization of the success of game player activity inWorld of Warcraft; a visualization of 311 number adoption that shows the diffusion of non-emergency calls in the United States; a return on investment study for two decades of HIV/AIDS research funding by NIAID; and a map showing the impact of the HiveNYC Learning Network.Visual Insightswill be an essential resource on basic information visualization techniques for scholars in many fields, students, designers, or anyone who works with data.

    eISBN: 978-0-262-32023-8
    Subjects: Library Science, Technology

Table of Contents

  1. Front Matter
    (pp. i-vi)
  2. Table of Contents
    (pp. vii-vii)
  3. Note from the Authors
    (pp. viii-viii)
  4. Preface
    (pp. ix-x)
    Katy Börner
  5. Acknowledgments
    (pp. xi-xi)
  6. Chapter One Visualization Framework and Workflow Design
    (pp. 1-35)

    Welcome to the Information Age, where each one of us receives more information via tweets, emails, news, and other data streams each day than can humanly be processed in 24 hours; and anyone with an Internet connection has access to a majority of humankind’s knowledge. Our offices are filling up and our email inboxes are overflowing (see Figure 1.1, left). We urgently need more effective ways to make sense of this massive amount of data—to navigate and manage information, to identify collaborators and friends, or to notice patterns and trends (see Figure 1.1, right).

    This book teaches you how...

  7. Chapter Two “WHEN”: Temporal Data
    (pp. 37-73)

    Chapters 2–6 introduce different types of analysis and resulting visualizations that answer a specific type of question. This chapter aims to answer “WHEN” questions using temporal data, analyses, and visualizations. The main goal is to understand the temporal distribution of datasets; to identify growth rates, latency to peak times, or decay rates; to see patterns in time-series data, such as trends, seasonality, or bursts.

    Each theory part of the subsequent five chapters starts with a discussion of exemplary visualizations followed by an overview and definition of key terminology, and introduction of general workflows. This chapter also introduces burst detection,...

  8. Chapter Three “WHERE”: Geospatial Data
    (pp. 75-111)

    In this chapter, we explore geospatial data analysis and visualization, which originated in geography and cartography but are increasingly common in statistics, information visualization, and many other areas of science. The analyses aim to answer “WHERE” questions that use location information to identify their position or movement over geographic space. For example, we might be interested to know where major experts are located, how they are interlinked via collaborations (Figure 1.17), or what career trajectory they took. Intangible entities, for example, an idea for a new product, might be born at a certain institution but might travel locally or abroad...

  9. Chapter Four “WHAT”: Topical Data
    (pp. 113-141)

    In this chapter, we will discuss topical (also called textural, linguistic, or semantic) data analysis and visualization to answer “WHAT” questions. The termtopic analysisis used in a variety of ways, but for the purposes of this book it means extracting a set of unique words or word profiles and their frequencies to determine the topic coverage of a body of text. That is, we will be using texts (e.g., from article abstracts or grant titles) to identify major topics, their interrelations, and their evolution over time at different levels of analysis—micro to macro.

    Just like the previous...

  10. Chapter Five “WITH WHOM”: Tree Data
    (pp. 143-167)

    This chapter provides an introduction on how to use tree data to answer “WITH WHOM” questions. Tree datasets, such as directory structures, organizational hierarchies, branching processes, genealogies, or classification hierarchies are commonly organized and displayed using tree visualizations: for example, tree views, treemaps, or tree graphs.

    This section discusses exemplary visualizations, relevant terminology, and different approaches to visualize hierarchical data.

    Here we will discuss four examples of tree visualizations. The first example was designed by Moritz Stefaner. It depicts a classification taxonomy developed and used in the Metadata for Architectural Contents in Europe (MACE) project (Figure 5.1).¹ This project aims...

  11. Chapter Six “WITH WHOM”: Network Data
    (pp. 169-213)

    In this chapter, we will look at the analysis and visualization of network data. The study of networks aims to increase our understanding of natural and manmade networks. It builds on social network analysis, physics, information science, bibliometrics, scientometrics, econometrics, informetrics, webometrics, communication theory, sociology of science, and several other disciplines.1,2,3Networks might represent collaborations between authors, business and marriage ties between families, or citation linkages between papers or patents. The goal of network studies is to identify highly connected authors (or papers), i.e., those with many collaboration (or citation) links; network properties such as size and density; structures such...

  12. Chapter Seven Dynamic Visualizations and Deployment
    (pp. 215-233)

    Some visualizations are too large or too complex to comprehend easily. In these cases, a dynamic deployment that supports interactive search, filtering, clustering, zoom and pan, or details on demand is beneficial. Tools like Tableau,¹ Gephi,² but also GUESS,³ and Cytoscape⁴ available as Sci2 Tool plugins, support interactive visualizations for data exploration and communication. In this chapter, we discuss the design of dynamic visualizations and the deployment of interactive visualizations via desktop programs, interactive online visualizations, and large touchscreens. Online visualization services such as Microscoft’s Zoom. it⁵ or Gigapan⁶ that support sharing of very high-resolution images such as largescale visualizations...

  13. Chapter Eight Case Studies
    (pp. 235-271)

    These case studies are the results of the Information Visualization MOOC 2013 client projects. The students were asked to form groups of four to five and select a real-world project from a list of potential client projects. The clients made their data available to the students, who worked to conduct a requirement analysis, develop an early sketch, conduct a literature review, preprocess and clean the data, and finally perform analysis and visualization. The students then submitted their visualizations to the client for validation. The following chapter highlights six of these projects, including feedback and insights from the clients.

    Local governments...

  14. Chapter Nine Discussion and Outlook
    (pp. 273-283)

    This chapter reviews lessons learned in the IVMOOC as delivered in Spring 2013. We briefly review feedback provided by students, present results of an in-depth analysis of IVMOOC data, and conclude with a discussion of planned activities related to the further development of MOOC content and delivery.

    The IVMOOC at Indiana University attracted 1,901 students from more than 90 countries. However, few of the IVMOOC 2013 students completed the course—a feature it shares with most other MOOCs. As a graduate-level course, the midterm and final were particularly demanding, and many students could not complete the client work due to...

  15. Appendix
    (pp. 284-291)
  16. Image Credits
    (pp. 292-293)
  17. Index
    (pp. 294-297)