Geographic Data Imperfection 1 E-Book

• Herausgeber: John Wiley & Sons
• Kategorie: Wissenschaft und neue Technologien
• Sprache: Englisch

Seitenzahl: 297

Veröffentlichungsjahr: 2019

Beschreibung

Geomatics is a field of science that has been intimately intertwined with our daily lives for almost 30 years, to the point where we often forget all the challenges it entails. Who does not have a navigation application on their phone or regularly engage with geolocated data? What is more, in the coming decades, the accumulation of geo-referenced data is expected to increase significantly. This book focuses on the notion of the imperfection of geographic data, an important topic in geomatics. It is essential to be able to define and represent the imperfections that are encountered in geographical data. Ignoring these imperfections can lead to many risks, for example in the use of maps which may be rendered inaccurate. It is, therefore, essential to know how to model and treat the different categories of imperfection. A better awareness of these imperfections will improve the analysis and the use of this type of data.

Leseprobe

Table of Contents

Cover

Preface

Part 1: Bases and Concepts

1 Imperfection and Geographic Information

1.1. Context

1.2. Concepts, representation, reasoning system, and data processing

1.3. Some conclusive remarks

1.4. References

2 Imperfection of Geographic Information: Concepts and Terminologies

2.1. Introduction

2.2. Semantics according to Humpty Dumpty

2.3. Taxonomies of GI and its related uncertainty

2.4. A theoretical framework of the nature of uncertainty and quality

2.5. Conclusion

2.6. References

3 The Origins of Imperfection in Geographic Data

3.1. Introduction

3.2. Imperfection during the life cycle of geographic data

3.3. The sources of the imperfections in a process

3.4. Examples of sources of imperfection in different processes

3.5. Conclusion

3.6. References

4 Integrity and Trust of Geographic Information

4.1. Introduction

4.2. The notions of quality

4.3. Internal quality and integrity

4.4. External quality and trust

4.5. Applying these notions to maritime geolocation data

4.6. Conclusion

4.7. References

Part 2: Representation

5 Formalisms and Representations of Imperfect Geographic Objects

5.1. Theories about the representation of an imperfect geographic object

5.2. Where and when do we refer to imperfection in geographic information?

5.3. Formalisms

5.4. Spatial objects

5.5. Reconsidering the introductory examples

5.6. References

6 Representing Diagrams of Imperfect Geographic Objects

6.1. Introduction

6.2. Describing the theoretical models of geographic objects

6.3. Describing the theoretical models of imperfect geographic objects

6.4. Toward massive databases

6.5. References

Part 3: Reasoning and Treatment

7 Algebraic Reasoning for Uncertain Data

7.1. Introduction

7.2. Algebras used for spatial reasoning

7.3. Lattices of relation

7.4. Extending these models to fuzzy regions

7.5. References

8 Reasoning in Modal Logic for Uncertain Data

8.1. Introduction

8.2. Reasoning in first-order predicate calculus

8.3. Reasoning in modal logic

8.4. References

9 Reviewing the Qualifiers of Imperfection in Geographic Information

9.1. Introduction

9.2. Belief revision and update in knowledge engineering

9.3. The limitations faced by GIS when representing a set of beliefs

9.4. Revision in a set of binary beliefs

9.5. The case of uncertain beliefs

9.6. Bayesian probabilistic conditioning

9.7. Revision in evidence theory

9.8. Possibilistic conditioning

9.9. Conclusion

9.10. References

10 The Features of Decision Aid and Analysis Processes in Geography: How to Grasp Complexity, Uncertainty, and Risks?

10.1. The decision-making context

10.2. Geographers, decision-makers, actors, and the territory

10.3. The objects, stakes, and issues involved in a decision

10.4. Information, data, knowledge, uncertainties, and bias

10.5. Supporting the structuring and resolution of ranking, choice, or sorting problems (issues)

10.6. A decision-analysis method for risk analysis and management

10.7. Conclusion

10.8. References

List of Authors

Index

End User License Agreement

List of Tables

Chapter 1

Table 1.1. Evolution of the appearance of the terms “spatial data” and “uncertai...

Chapter 2

Table 2.1. Different definitions of the word “error”

Chapter 4

Table 4.1. Position piracy alerts

Table 4.2. Alerts sent out because of dynamic data inconsistency

Chapter 5

Table 5.1. Examples of t-norms and t-conorms used in the frame of spatial object...

Chapter 6

Table 6.1. Objects that belong to the class Imperfection

Chapter 7

Table 7.1. A composition table for the algebra of points − for Allen’s algebra, ...

Table 7.2. The composition table for the basic relations in RCC8

Chapter 8

Table 8.1. Sample facts about spreading

Table 8.2. The set of formulae for spreading

Table 8.3. Facts related to the spreadable products

Table 8.4. Results obtained by solving the goal ◊ spread_on(P, Z)

Table 8.5. Rules of the field

Table 8.6. Results of ◊ detected_spreading(P, PA) with different axioms

Chapter 9

Table 9.1.

Table 9.2.

Table 9.3.

Table 9.4.

Table 9.5.

Table 9.6.

Table 9.7.

Table 9.8.

Table 9.9.

Table 9.10.

Table 9.11.

Table 9.12.

Guide

Cover

Table of Contents

Begin Reading

Pages

v

iii

iv

xi

xii

1

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

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

105

106

107

108

109

110

111

112

113

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

175

176

177

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

Geographic Data Imperfection 1

From Theory to Applications

Edited by

Mireille Batton-Hubert

Eric Desjardin

François Pinet

First published 2019 in Great Britain and the United States by ISTE Ltd and John Wiley & Sons, Inc.

Apart from any fair dealing for the purposes of research or private study, or criticism or review, as permitted under the Copyright, Designs and Patents Act 1988, this publication may only be reproduced, stored or transmitted, in any form or by any means, with the prior permission in writing of the publishers, or in the case of reprographic reproduction in accordance with the terms and licenses issued by the CLA. Enquiries concerning reproduction outside these terms should be sent to the publishers at the undermentioned address:

ISTE Ltd

27-37 St George’s Road

London SW19 4EU

UK

www.iste.co.uk

John Wiley & Sons, Inc.

111 River Street

Hoboken, NJ 07030

USA

www.wiley.com

© ISTE Ltd 2019

The rights of Mireille Batton-Hubert, Eric Desjardin and François Pinet to be identified as the authors of this work have been asserted by them in accordance with the Copyright, Designs and Patents Act 1988.

Library of Congress Control Number: 2019938868

British Library Cataloguing-in-Publication Data

A CIP record for this book is available from the British Library

ISBN 978-1-78630-297-7

Preface

Geomatics is a scientific field that in the last 30 years has become closely intwined with our everyday life, to such an extent that we often forget all its underlying challenges. Who does not have a navigation application on his or her mobile phone? Who does not manipulate geolocated data? In the coming decades, the volumes of georeferenced data generated should increase dramatically.

This book focuses on the notion of imperfection in geographic data, which is a significant topic in geomatics. In fact, it is essential to define and represent the imperfection that may affect geographic data. Uncertainty constitutes the basis of the study of the so-called modern probability, a field that became very active in the 18th century (thanks to the works carried out by P. de Fermat, B. Pascal, Th. Bayes, P. S. Laplace, and several others) and was complemented by concepts developed in the 19th century and, more particularly, later in the 20th century. The notion of imperfection supplements this concept; the single representation of the stochastic (random) nature of a fact is limited when the aim is to represent the precision of a fact and/or the lack of knowledge about data. These theories, which deal with these two aspects, were complemented in particular by the Dempster–Shafer theory.

A better awareness of this imperfection, which is linked specifically to geographic data, especially during the formalization, storage, and manipulation of this characteristic, improves their analyses and any decision-analysis process. Even if it is important (and even critical) to manage imperfection, it must be acknowledged that integrating it into data-processing procedures may be challenging. To take up this challenge, this book intends to bridge the gap between the need and its implementation. It simultaneously explores theoretical aspects, in order to illustrate more clearly phenomena and representations, and practical/pragmatic aspects by presenting concrete examples and applied tools.

This book was written in the context of an initiative of the Groupement de Recherche du CNRS sur les Méthodes et Applications pour la Géomatique et l’Information Spatiale (GDR MAGIS) (Associated Research of the CNRS on the Methods and Uses of Geomatics and Spatial Information). This initiative, which targeted the uncertainty of spatial data, gave rise to a specific work group which took part in writing this book. Thus, this book is the common product of an analysis of this topic. It is our hope that it will manage to meet the readers’ expectations.

We would like to express our sincere thanks to the authors of the various chapters and, more generally, to all the individuals who took part in the work groups of the GDR MAGIS over time. We extend our thanks to them for their fruitful ideas, which have made it possible to elaborate further on the ideas expressed in this book. We would like to thank the GDR MAGIS of the CNRS as well as its various directors for their support and their trust in this project.

We hope that readers will enjoy this book and that it will shed some light on the methods that make it possible to better understand and process geographic imperfections.

The editors of this book and the organizers of the initiative Incertitude épistémique – des données aux modèles en géomatique (Epistemic Uncertainty – from Data to Models in Geomatics) of the GDR MAGIS of the CNRS:

Mireille BATTON-HUBERT

Eric DESJARDIN

François PINET

May 2019

Part 1Bases and Concepts

2Imperfection of Geographic Information: Concepts and Terminologies

2.1. Introduction

Geographic information results from a process, which is often complex, that implies collecting information about the so-called “real” world, abstracting and simplifying it so that it can be represented in an environment which is in general digital. A process of this kind cannot, and should not, capture the world in all its complexity. Some of these induced/generated imperfections are accidental (like errors created during the collection of data), whereas others may have been introduced on purpose (cartographic generalization) to provide a simplified view of the world that corresponds to the needs of a specific community of users. The sources and nature of these imperfections have been relatively well-known, described, and assessed/qualified/measured for a long time in cartography. This body of knowledge was integrated in the 1980s in Geographic Information Sciences (GISciences) and geomatics, as well as in various other fields (see, e.g. [BÉD 86, CHR 83, GOO 83, ROB 85] for discussions of these questions). Often identified as “errors” in early works [CHR 91, FIS 87, GOO 89], these imperfections were later described using a wider range of terms, which led them later on to fall under the two broad concepts of spatial uncertainty and data quality. There are several words employed to describe the differences between the world and its representation, and, depending on their nature or source, they are often associated with terms like accuracy, ambiguity, completeness, consistency, error, ignorance, imperfection, precision, quality, random nature, uncertainty, and vague/fuzzy. Even if a consensus has been reached on the reasons behind the source of these imperfections, there are strongly divergent opinions about the terms used to describe each of these types, the exact meaning of these terms, and the way in which they interfere/interact with one another. Various authors have defined and even organized these terms within ontologies [BÉD 86, DUC 01, FIS 99, SMI 89], thus expressing a point of view on their definitions and relationships, but without necessarily looking for an agreement within the community on the use of a common terminology. Such a lack of consensus about a common language has been criticized in the past and was considered to negatively affect some of the research carried out by the various communities interested in geographic information (see, e.g. [DEV 10]).

Fisher [FIS 03], for instance, has criticized the general lack of links between the fields that study spatial uncertainty and spatial data quality. This argument was developed later on by [COM 06]. Fisher uses the analogy of “ships passing in the night”, where two ships (i.e. the fields of “uncertainty” and “data quality”) may be quite close in space without necessarily seeing each another).

Here, we aim to shed some light on this topic, potentially helping these ships and perhaps others to see each other more clearly and understand what makes them similar or different.

This introductory chapter does not aim to provide a single authoritative terminology, or to create a new taxonomy, but, on the contrary, to acknowledge the various ways in which these terms are used in different fields. We will talk about these differences and similarities through the use of these terms by associating them with certain causes that are at the root of these imperfections and the way they are used.

Section 2.2 discusses some terms used to describe spatial uncertainty, data quality, and their interrelations. Section 2.3 presents several taxonomies of spatial uncertainty or related concepts. Ultimately, section 2.4 presents a theoretical/conceptual framework that discusses the nature of the uncertainty and data quality, as well as the elements that they share or in which they differ.

2.2. Semantics according to Humpty Dumpty1

Uncertainties are inherent to various descriptions of our environment. When we describe our world, its large size and complexity force us to merely observe some of its parts (e.g. by sampling). Our representations are also necessarily simplified (e.g. by summarizing the information through descriptive statistics, by carrying out map generalization processes, or by combining information in sub-divisions of space such as pixels). Descriptions of our world, whether written, oral, or graphical, like books, paintings, or speeches, provide a piece of information about our environment which is, on different levels, incomplete and inaccurate. [BÉD 86] described this process very accurately in a modeling exercise necessary for the mapping communication process.

In common language, some words or tenses (e.g. the conditional) can be employed to characterize the level of certainty assigned to information and sometimes even suggest acceptable probability levels (e.g. [TEI 88]). For instance, it could be said that the economic crisis “might” continue next year, that the weather “will quite likely” be good tomorrow, or that we are “near” the end. In common language, a large number of words can also be used to describe the nature of uncertainties or imperfections. For example, information may be vague or ambiguous, an image may be blurry, a forecast may be accurate, and so on. While several languages have a rich vocabulary that can be used to represent these aspects, the meaning of these terms may be vague or unknown to most individuals. The definitions of these terms as they are given in dictionaries may not be very helpful, as a definition is often established by employing other related terms, often giving the impression that many of these terms can be used as synonyms (Figure 2.1). Some studies have also shown that similar words that describe uncertainty do not necessarily mean the same thing for individuals who speak different languages [DOU 03].

While some of these words may have similar meanings, others have clearly distinct meanings; for example, “accuracy” is perceived to be more similar to “precision” than “incompleteness” or “incoherence”.

Figure 2.1.Relations between different words found in the Webster English dictionary. Arrows link words (origin) whose definitions refer to other words (the arrowhead). Full arrows indicate positive relations, while the dotted arrows indicate negative relations (namely, when a word is defined by negating another word)

While some words may mean the same thing for most individuals (e.g. the words “precise” and “accurate” are often regarded as synonymous in common parlance), other words may mean something different depending on the communities of individuals. Therefore, words can mean different things to individuals based on their experience or areas of expertise. People occasionally adopt intolerant behaviors toward those who use definitions that differ from theirs. This is not dissimilar from a comment made by Humpty Dumpty, a popular character from literature who appears in Lewis Carroll’s Through the Looking Glass, and What Alice Found There [CAR 71]. Humpty Dumpty (Figure 2.2) is discussing semantics with Alice when he says: “When I use a word […] it means just what I choose it to mean – neither more nor less”. Similarly, several individuals or groups associate their own meanings to words, neither more nor less, sometimes adopting specific definitions for some terms that could have different meanings for other communities.

Figure 2.2.An excerpt from Through the Looking Glass, and What Alice Found There [CAR 71]

Diversity in the words used to describe uncertainty and imperfections in everyday speech can also be found in science. Subjects like statistics, economics, or medicine may assign different meanings to a same word or use two different words to refer to a same concept. This diversity can also be found in geomatics, a highly interdisciplinary field. For example, “uncertainty” and “error” are synonyms for some authors, whereas they mean different things for others. Authors can also assign different meanings to the same word. Table 2.1