The Fitness of Information E-Book

Chaomei Chen

• Herausgeber: John Wiley & Sons
• Kategorie: Wissenschaft und neue Technologien
• Serie: Wiley Series in Probability and Statistics
• Sprache: Englisch
• Veröffentlichungsjahr: 2014

Seitenzahl: 542

Beschreibung

Theories and practices to assess critical information in a complex adaptive system Organized for readers to follow along easily, The Fitness of Information: Quantitative Assessments of Critical Evidence provides a structured outline of the key challenges in assessing crucial information in a complex adaptive system. Illustrating a variety of computational and explanatory challenges, the book demonstrates principles and practical implications of exploring and assessing The Fitness of Information in an extensible framework of adaptive landscapes. The book's first three chapters introduce fundamental principles and practical examples in connection to the nature of aesthetics, mental models, and the subjectivity of evidence. In particular, the underlying question is how these issues can be addressed quantitatively, not only computationally but also explanatorily. The next chapter illustrates how one can reduce the level of complexity in understanding the structure and dynamics of scientific knowledge through the design and use of the CiteSpace system for visualizing and analyzing emerging trends in scientific literature. The following two chapters explain the concepts of structural variation and The Fitness of Information in a framework that builds on the idea of fitness landscape originally introduced to study population evolution. The final chapter presents a dual-map overlay technique and demonstrates how it supports a variety of analytic tasks for a new type of portfolio analysis. The Fitness of Information: Quantitative Assessments of Critical Evidence also features: * In-depth case studies and examples that characterize far-reaching concepts, illustrate underlying principles, and demonstrate profound challenges and complexities at various levels of analytic reasoning * Wide-ranging topics that underline the common theme, from the subjectivity of evidence in criminal trials to detecting early signs of critical transitions and mechanisms behind radical patents * An extensible and unifying framework for visual analytics by transforming analytic reasoning tasks to the assessment of critical evidence The Fitness of Information: Quantitative Assessments of Critical Evidence is a suitable reference for researchers, analysts, and practitioners who are interested in analyzing evidence and making decisions with incomplete, uncertain, and even conflicting information. The book is also an excellent textbook for upper-undergraduate and graduate-level courses on visual analytics, information visualization, and business analytics and decision support systems.

Leseprobe

CONTENTS

Cover

Wiley Series in Probability and Statistics

Title page

Copyright page

Preface

Chapter 1: Attention and Aesthetics

1.1 Attention

1.2 Gestalt Principles

1.3 Aesthetics

1.4 The Index of the Interesting

1.5 Summary

Bibliography

Chapter 2: Mental Models

2.1 Mental Models

2.2 Creativity

2.3 Foresights

2.4 Summary

Bibliography

Chapter 3: Subjectivity of Evidence

3.1 The Value of Information

3.2 Causes Célèbre

3.3 The Da Vinci Code

3.4 Supreme Court Opinions

3.5 Apple versus Samsung

3.6 Summary

Bibliography

Chapter 4: Visualizing the Growth of Knowledge

4.1 Progressive Knowledge Domain Visualization

4.2 CiteSpace

4.3 Examples

4.4 Summary

Bibliography

Chapter 5: Fitness Landscapes

5.1 Cognitive Maps

5.2 Fitness Landscapes

5.3 Applications of Fitness Landscapes

5.4 Summary

Bibliography

Chapter 6: Structural Variation

6.1 Complex Adaptive Systems

6.2 Radical Patents

6.3 Bridging the Gaps

6.4 Applications

6.5 Summary

Bibliography

Chapter 7: Gap Analytics

7.1 Portfolio Analysis and Risk Assessment

7.2 Interactive Overlays

7.3 Examples of Dual-Map Overlays

7.4 Summary

7.5 Conclusion

Bibliography

Index

Wiley Series in Probability and Statistics

End User License Agreement

List of Tables

Chapter 01

Table 1.1 The ratios between adjacent Fibonacci numbers approaching the golden ratio

Table 1.2 Dialectical relations between taken-for-granted assumptions and new propositions

Chapter 03

Table 3.1 Statistics of the reviews

Table 3.2 Statistics of the set of U.S. Supreme Court Opinions

Table 3.3 Top 100 most cited cases in the dataset

Chapter 04

Table 4.1 Variables associated with articles that may be predictive of their subsequent citations

Table 4.2 Users from different cities may study the same topics

Chapter 06

Table 6.1 The first two levels of the dewey decimal system

a

Table 6.2 The third level sections under science (500) in the dewey classification system

Table 6.3 Zinb regression and nb models of global citation counts of 3515 citing articles on complex network analysis (1996–2004)

Table 6.4 Zinb model of 2065 patents in the expanded Nci patents dataset

Table 6.5 NB model of the patent citations

Table 6.6 The ZINB model with more independent variables

Chapter 07

Table 7.1 Awards labeled in Figure 7.3

Table 7.2 Portfolios of three organizations’ publications during 2008 and 2012

Table 7.3 Journals involved in the regenerative medicine dataset (2005–2012)

Table 7.4 The number of journals involved in articles citing

JASIST

Table 7.5 Citing trajectories at the discipline level

List of Illustrations

Chapter 01

Figure 1.1 A preference of the closed circle.

Figure 1.2 Are you seeing a circle?

Figure 1.3 Continuity.

Figure 1.4 The golden ratio calculation.

Figure 1.5 The drawing on the left consists of segments from various sized circles.

Figure 1.6 Photographs are combined to identify common features.

Figure 1.7 Information entropies of the literature of terrorism research between 1990 and the first half of 2007. The two steep increases correspond to the Oklahoma City bombing in 1995 and the September 11 terrorist attacks in 2001.

Figure 1.8 Symmetric relative entropy matrix shows the divergence between the overall use of terms across different years. The recent few years are most similar to each other. The boundaries between areas in different colors indicate significant changes of underlying topics.

Figure 1.9 A network of keywords in the terrorism research literature (1995–1997). High-frequency terms are shown in boldface, whereas outlier terms identified by informational bias are shown in italic.

Chapter 02

Figure 2.1 Mental models are easy to form but hard to change.

Figure 2.2 Connecting the dot with no more than four jointed straight lines.

Figure 2.3 The three clusters of cocited papers can be seen as three patches of information. All three patches are about terrorism research. Prominently labeled papers in each patch offer information scent of the patch. The sizes of citation rings provide a scent of citation popularity.

Chapter 03

Figure 3.1 The tip of an iceberg and the shape of the iceberg as a whole.

Figure 3.2 A sketch of a Wigmore chart.

Figure 3.3 Prosecution’s top-level propositions in the Sacco and Vanzetti case.

Figure 3.4 Coherent and incoherent explanations and evidence.

Figure 3.5 “O.J. Simpson killed Nicole” has more incoherent links than other propositions.

Figure 3.6 The distribution of customer reviews of

The Da Vinci Code

on Amazon.com within the first year of its publication (March 18, 2003–March 30, 2004). Although positive reviews consistently outnumbered negative ones, arguments and reasons behind these reviews are not apparent.

Figure 3.7 A decision tree representation of terms that are likely to differentiate positive reviews from negative reviews made in 2003.

Figure 3.8 A decision tree based on reviews made in 2004.

Figure 3.9 An opinion differentiation tree of Video iPod reviews.

Figure 3.10 Representing emerging topics in 3944 abstracts of publications on terrorism.

Figure 3.11 Examples of citations to the case

Miranda v. Arizona

, 384 U.S. 436 (1966).

Figure 3.12 Top 20 cases with the strongest citation bursts.

Figure 3.13 A visualization of a cocitation network of cases cited with the case

Miranda v. Arizona

, 384 U.S. 436 (1966). The size of a circle is proportional to the citations received by the case. Red circles represent cases with citation bursts.

Chapter 04

Figure 4.1 An ultimate ability to reduce the vast volume of scientific knowledge in the past and a stream of new knowledge to a clear and precise representation of a conceptual structure: (1) domain knowledge prior to Galea et al. (2002), (2) new discovery reported in Galea et al. (2002), and (3) updated domain knowledge.

Figure 4.2 An animated visualization of a network-based landscape view of the literature on mad cow disease.

Figure 4.3 The relationship between a research front and its intellectual base.

Figure 4.4 Three types of salient nodes in a cocitation network.

Figure 4.5 An architecture of the CiteSpace system.

Figure 4.6 The procedure of generating local maps in CiteSpace.

Figure 4.7 The procedure of a multiple-perspective analysis.

Figure 4.8 The procedure of predictive analysis of scientific literature.

Figure 4.9 A segment of a sequence of interactive events of a user conducting a document cocitation analysis in CiteSpace.

Figure 4.10 A segment of a sequence of interactive events as a user exploring various control functions in CiteSpace.

Figure 4.11 Major areas in terrorism research.

Figure 4.12 The simplicity of the research field at the cluster level.

Figure 4.13 A visualization of a 3638-node network generated by sampling top 500 most cited articles each year between 1996 and 2003.

Figure 4.14 A network of 12,691 cocited references generated by sampling top 2000 most cited articles per year.

Figure 4.15 Trends in mass extinctions research.

Figure 4.16 The patterns identified in our 2006 article were found 4 years later by experts of mass extinctions in 2010.

Figure 4.17 A visualization of mass extinction in a follow-up study in 2011.

Figure 4.18 A timeline visualization of mass extinctions research.

Figure 4.19 The distribution of CiteSpace’s users. The height of a bar represents the average size of networks analyzed by users at a city.

Figure 4.20 The topics of the largest clusters in networks analyzed by CiteSpace users grouped by IP locations.

Figure 4.21 A visualization of U.S. Supreme Court landmark cases.

Chapter 05

Figure 5.1 Sewall Wright’s diagrams of adaptive landscapes. Each frame represents different evolutionary scenarios and their impact on the population. Frame C represents a dynamic landscape in a changing environment. Frame F represents the dynamics of Wright’s shifting balance theory.

Figure 5.2 Swiss Cheese Model of system failures.

Figure 5.3 The evolution of the AA2 dataset is shown in four snapshots representing the structure–activity landscape in 2001, 2003, 2005, and 2007, respectively.

Figure 5.4 A close-up view of three-dimensional SAR landscape of the AA2 dataset in 2005.

Figure 5.5 Skandia’s IC landscape. Am, American Skandia; Di, Dial; In, Intercaser; UK, UK Life.

Figure 5.6 Top-down view of Skandia’s performance on an IC-fitness landscape.

Figure 5.7 Fitness landscape of the life expectancy at birth in 44 countries from 1970 to 1995. Lighter colors represent longer life expectancy.

Figure 5.8 A rising landscape of research about mad cow disease (green), CJD in human (blue), and the fundamental research on prions (red).

Figure 5.9 Thematic landscapes of computer graphics (1982–1999).

Chapter 06

Figure 6.1 A newly found connection changes the existing structure of the domain knowledge of PTSD.

Figure 6.2 Signals may cause critical transitions in a CAS of knowledge.

Figure 6.3 Incremental changes (a) and transformative changes (b) caused by a perturbation to the system.

Figure 6.4 Boundary-spanning mechanisms modeled in the structural variation theory.

Figure 6.5 Modularity measures changed in a series of cocitation networks of regenerative medicine.

Figure 6.6 The synthesized network of research in small-world network.

Figure 6.7 A network of cocited references derived from 5135 articles published on small-world networks between 1990 and 2010. The network of 205 references and 1164 cocitation links is divided into 12 clusters with a modularity of 0.6537 and the mean silhouette of 0.811. The red lines are made by the top-15 articles with the largest centrality variation rate.

Figure 6.8 A Voronoi diagram of patents by IPC. Red, IBM; green, Google; yellow, both.

Figure 6.9 A minimum spanning tree of a network of 1726 cocited patents related to cancer research.

Figure 6.10 A timeline visualization of a broader context of the NCI patents. Dashed lines represent transformative links connecting different clusters of patents. Patents in each cluster are shown horizontally by their granted date from left to right. Labels next to clusters on the right characterize the primary topics of impact.

Figure 6.11 The impact of patent 4676980 on the existing network structure. Before patent 4676980 was granted, patents A and C were not cocited at the patent level. After it was granted, patents A and C became cocited with a strength of 0.035.

Figure 6.12 A patent may introduce structural changes at one level of granularity but not at other levels. Here patent 4676980 connected patents A and C, which were not connected at the patent level, but it didn’t introduce any structural changes at the IPC level because it was assigned to the group of IPC classes that belong to the triangle. At the U.S. classification level, patents A, B, and C’s classes are already connected.

Figure 6.13 U.S. Patent 6537746, which was identified with the highest

C

KL

.

Figure 6.14 A timeline visualization of cocited patents. The star on the top is U.S. Patent 6537746. It contributed novel links connecting clusters from 83 through 88.

Chapter 07

Figure 7.1 The procedure for identifying core passages of a full-length document.

Figure 7.2 An illustration of the distribution of transformative research along two dimensions.

Figure 7.3 Transformative potentials of awards.

Figure 7.4 Transformative potentials of proposals (+: awarded; −: declined). Size = the amount requested.

Figure 7.5 An overview illustrates the construction and use of dual-map overlays. Citation arcs, cocitation links, and trajectories over time facilitate the study of multiple sets of publications at an interdisciplinary level, an organizational level, and the individual publication level.

Figure 7.6 The initial appearance of the Dual-Map user interface, showing both citing and cited journal base maps simultaneously. The base map of 10,330 citing journals is on the left. The base map of 10,253 cited journals is on the right. The colors depict clusters identified by the Blondel clustering algorithm.

Figure 7.7 The boundary of each cluster is shown to depict how its members are distributed. Clusters in both base maps overlap substantially.

Figure 7.8 Citation patterns in an overlay of 405 articles that cited the Wakefield paper.

Figure 7.9 Overlays of three iSchools show major threads of citations that may characterize the publication portfolios of these institutions. The lower half of the figure shows the citing and cited trajectories in each of the base maps.

Figure 7.10 Trajectories of Google (blue), Microsoft (red), and IBM (green).

Figure 7.11 Citation overlays of three corporations.

Figure 7.12 Trajectories of regenerative medicine research (2005–2012). The citing trajectory remains to be in the disciplinary area labeled as molecular, biology, and immunology throughout the entire course.

Figure 7.13 Trajectories of research in mass extinctions (1975–2010) at the discipline level. The core discipline of the research is identified as the Blondel cluster 3 on ecology, earth, and marine. The longest single-year shift occurred between 1978 and 1979 as the disciplinary center of the journals shifted from the Blondel cluster 5 on physics, materials, and chemistry to the Blondel cluster 4 on molecular, biology, and immunology.

Figure 7.14 An overlay of publications in visual analytics (2006–2012). Wavelike curves depict citation links. They are colored by their source clusters. Dashed lines depict cocitation links across disciplinary boundaries.

Figure 7.15 An overlay of articles citing

JASIST

(2002–2011).

Figure 7.16 Characteristics of trajectories of Witten’s publications and Shneiderman’s publications. Citing trajectories of overlays are shown on the left. Cited trajectories are shown on the right.

Figure 7.17 A fitness landscape of scientific inquires.

Guide

Cover

Table of Contents

Begin Reading

Pages

ii

iii

iv

ix

x

xi

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

266

267