See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/291335624

A Guide to Geospatial Data Quality

Presentation · June 2015

DOI: 10.13140/RG.2.1.2819.1763

CITATIONS READS

3 5,966

3 authors, including:

Yvan Bédard Sonia Rivest

Yvan Bedard Photo Art Laval University
250 PUBLICATIONS 3,306 CITATIONS 29 PUBLICATIONS 781 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

3D Modeling and Data Quality View project

Fundamental definitions in Geomatics and Geospatial Data View project

All content following this page was uploaded by Yvan Bédard on 21 January 2016.

The user has requested enhancement of the downloaded file.

 Earth Sciences Sector

Welcome to
the Webinar!

Eric Wright Yvan Bédard

A Guide to Geospatial Data Quality

Thank you for waiting while other participants

Teleconference ID: join the teleconference and the Webinar.
748 918 8 We’ll be starting in a few minutes.

To join the teleconference:

Toll-free in Canada & US: 1-877-413-4790
Ottawa and International: 1-613-960-7514
 2

Goal: This webinar will inform you on the importance

of geospatial data quality, the concepts underlying
geospatial data quality, the fundamentals of
geospatial data quality evaluation and risk
management, geospatial data quality evaluation and
risk management in practice, and recommendations in
B2B, B2C and C2C contexts.

Reminder
Please put your mobile devices on
mute during the presentation.
 3

Other registrants:
CGDI Webinar: Guide to Geospatial Data Quality • Netherlands
• Sweden
• Ireland
• Spain
• USA
• Ukraine
• Nigeria
• Uruguay
• Ecuador
• Spain

1 1

10

2 2
5 6
58
4

Registrants in Canada: 92

Registrants in Total : 113

 4

Today’s program:

 13:30 Welcome and Introduction

 13:50 Guide to Geospatial Data Quality

 14:40 Questions and Answers (via Webex chat)

 14:55 Summary and Conclusions

Introduction to GeoConnections
 6

GeoConnections
The GeoConnections program is a national initiative, led by Natural
Resources Canada, designed to facilitate access to and use of authoritative
geospatial information in Canada. GeoConnections supports the integration
and use of the Canadian Geospatial Data Infrastructure (CGDI).

Key Program Activities:

Geospatial Strategy and Leadership – continued coordination of

geomatics activities in Canada, requiring the development and
implementation of long-term national geomatics strategies and policies,
in partnership with CGDI stakeholders.

Canadian Geospatial Data Infrastructure – work with the geomatics

community to advance the operational policies and standards needed
to complete the CGDI and support the use of geospatial information.
 7

GeoConnections: Objectives

 Create increased awareness of the benefits of using geospatial

data and tools to achieve goals for social, economic and
environmental priorities.

 Facilitate the integration and use of geospatial data to support

effective decision making.

 Coordinate the development of national policies, standards and

mechanisms and support their implementation to ensure
maintenance and updating of geospatial data and compatibility with
global standards.

 Keep Canada at the leading edge of accessing, sharing and using

geospatial information via the Internet.
The Canadian Geospatial Data
Infrastructure
 9

What is the CGDI?

CGDI Components and
Guiding Principles
 The CGDI is an online network of
resources that improves the sharing, use
and integration of information tied to
geographic locations in Canada.
 In essence, via collaboration, the CGDI is
the convergence of policies, standards,
technologies, and framework data
necessary to harmonize all of Canada’s
location-based information.
 Through the CGDI, Canadians can discover,
access, visualize, integrate, apply and share
quality location-based information. The CGDI
allows citizens to gain new perspectives
into social, economic, and environmental
issues and make effective decisions.
CGDI – Overview; CGDI Vision, Mission and Roadmap:
http://geoconnections.nrcan.gc.ca/18
 10

Collaboration and Interoperability

 Collaboration, partnerships and a common way forward between
federal, provincial, territorial and regional governments; the private
sector; and academia ensure interoperability for the CGDI.

 Interoperability is achieved by the convergence of framework data,

policies, standards and technologies necessary to harmonize
Canada’s location-based information.
11
 12

Guide to Geospatial Data Quality

 Discusses why geospatial data quality is important

 Reviews the concepts underlying geospatial data
quality
 Discusses the geospatial data quality evaluation
process in details (based on ISO 19157 and ISO
19158)
 Discusses the management of risks of inappropriate
use of geospatial data (based on ISO 31000)
 Presents detailed examples of quality evaluation and
risk management tasks to be undertaken in the B2B,
B2C and C2C contexts
 13

Guest Speaker
Dr. Yvan Bédard is a consultant in research management and the
Strategic/Scientific Senior Advisor of Intelli3, a private company
specializing in Geo-BI and GeoData Analysis. He was a full professor
and successful researcher for 28 years at Université Laval Department
of Geomatics Sciences. He was the founding-director of the Centre for
Research in Geomatics and a member of several strategic and scientific
committees in Canada. His expertise is in GIS/Spatial
Databases/Business Intelligence and Analytics as well as in the most
recent spatial data quality issues. He holds a B.Sc.A. in Surveying, a
M.Sc. in Geodesy, and a Ph.D. in Civil Engineering while he's been
involved in computer sciences for over 35 years.
A Guide to Geospatial Data Quality
 15

Guide Outline

 Introduction: Why is geospatial data quality important?

 Background: the concepts underlying geospatial data
quality
 Geospatial data quality evaluation and risk management:
the fundamentals
 Geospatial data quality evaluation and risk management in
practice
 Recommendations for B2B, B2C and C2C contexts
 Conclusions
 16

Introduction
Section Overview

 The new context

 Geospatial data quality:
 Why?
 Who should care?
 Geospatial data quality and risk management
 Why this guide?
 17

The new context

 The production, distribution and usage of geospatial data

have changed dramatically in the last decade: new players

 Citizens are massively consuming and producing geospatial

data:
 Smartphones
 Navigation systems
 Numerous sources of digital maps and imagery
 Virtual globes, …

 It is the beginning of a new era…

(…) 1950 1960 1970 1980 1990 2000 2000 2010 2020 (…)

Hardware Era Software Era Data Era

 18

The new context

 The context is evolving…
B2B

B2B B2C

C2C
 The roles and responsibilities of producers, distributors and
users of geospatial data and services are evolving

 National geospatial data infrastructures are adapting to the

new challenges of data quality evaluation, quality
information communication and risk management
 19

Why is geospatial data quality important ?

 Data are collected to meet well-defined goals and their
quality is typically defined to meet these goals

 To minimize costs and delays, geospatial data reuse and

interoperability have become very popular
 But finding data that fit an intended purpose is not easy

 In the new B2C and C2C contexts, typical geospatial data

users do not understand the uncertain nature of geospatial
data and take digital data for granted:
 Erroneous decisions, possibly having significant social, political or
economic consequences
 Incidents of all sorts (injuries, material damages, 911 delays, people
lost in the wild, …)
 20

Why is geospatial data quality important ?

Death of a tourist: GPS map

error (©The Globe and Mail)
Plane crash and death of 47 Death of a lost skier:
passengers: wrong runway map completeness error,
missing path on map

Overlaying cadastral data

on aerial photo, he cuts the
bush fence and has a fight
22 injured students: erroneous with his neighbor, one Wrong house demolished:
vertical clearance person at the hospital GPS data error
 21

Why is geospatial data quality important ?

 Regulations and court decisions suggest that consumers
must be informed about the quality of a product or service,
as well as about the risks and prohibited usages:
 User manuals, guarantees, warnings, …

 A recent survey in the Canadian geomatics industry

indicated that (Gervais, Bédard, Larrivée, Rivest, & Roy, 2013):
 70% of respondents believe users are not aware of the potential
risks of using geospatial data
 Respondents show strong concerns about users’ ability to manage
risks
 81% thought that the geospatial industry could do more to reduce
users’ risks of inappropriate use of geospatial data
 Users’ primary complaints concerned poor documentation and data
quality
21
 22

Why is geospatial data quality important ?

 The CGDI aims to help the geospatial community meet the
data quality challenge by facilitating:
 The comparison of data sources
 The analysis of fitness-for-use
 The estimation of the cost of preparing data
 The evaluation of the risks of inappropriate data usages
 The communication about quality and risks to users

 Data quality is a key to good decision making

22
 23

Who should care about geospatial data quality ?

 In the B2B context, every player involved contributes to

define data quality, from needs analysis and database
design to the acquisition, integration, transformation and
dissemination of geospatial data

 Traditionally, B2B context (experts talk to experts):

 Completeness
 Logical consistency
 Positional accuracy Metadata
 Thematic accuracy
 Temporal quality

23
 24

Who should care about geospatial data quality ?

 Today, the new B2C and C2C contexts have emerged:

 Redistribution and reuse of data (known and unknown)
 Geospatial data and services have become a mass market, the
general rules of quality, information and responsibility-sharing apply

 In these new B2C and C2C contexts, who should care?

 Every designer and developer of systems using geospatial data
(including app developers for smartphones and web-based mapping
applications)
 Every provider of geospatial data (including VGI volunteers)
 Every expert involved in geoprocessing and disseminating
representations of geospatial data (including open-data projects)
 Every user of geospatial data

24
 25

Geospatial data quality and risk management

The Complete Cycle of Geospatial Data Quality
for a Spatially-Enabled Society Users
Producers
DESCRIBING CONTEXT

DEFINING NEEDS/REQUIREMENTS
(ISO 19131)
(ISO TC 145/ISO 3864-2)
PROPERLY WITH ALL

Selecting/Producing a dataset
COMMUNICATING

REVIEWING AS
MONITORING
NECESSARY
PLAYERS

EVALUATING DATA QUALITY

(ISO 19157)

dataset
accepted?
No
Yes
MANAGING RISK of
inappropriate use of geospatial
data (ISO 31000)

25
 26

Geospatial data quality and risk management

 Caring about geospatial data quality requires an evaluation of

this quality
 Depending upon the extent of the evaluation, it can be very
complex and robust or it can remain general and show a
higher level of uncertainty

In the B2B context (experts serve experts)

 Standardized geospatial data quality evaluation processes are
well described in the ISO 19157 IS
 27

Geospatial data quality and risk management

In the B2C context (experts serve non-experts)

 Professional duty to look at risk management strategies to
reduce potential negative impacts on non-expert users
 Such a duty emanates from the Code of Ethics of licensed
professionals, the concept of precaution, consumer protection
philosophy and liability issues

In the C2C context (non-experts serve non-experts)

 The concepts of precaution, consumer protection philosophy
and liability issues also call for risk management
 28

Geospatial data quality and risk management

 Risk management helps to select the best strategies to
minimize the risk of inappropriate use of geospatial data in
situations of uncertainty:
 Uncertain data
 Uncertain data quality
 Uncertain usages
 Uncertain expertise of users, ...
 Such strategies:
 Raise the awareness of users and providers
 Reduce the overall risks
 Help recognize the responsibility of every player
 Geospatial data quality results must be well communicated
to users in a language they understand AND actions
performed to minimize the risks related to geospatial data
usage
 29

Why this Guide ?

 Objective: support the Canadian geospatial community into

its efforts to make the spatially-enabled Society more aware
of geospatial data quality with the help of international
standards such as ISO 19157 (Geospatial Data Quality) and
ISO 31000 (Risk Management)

 Facilitate the selection of fit-for-purpose data

 Facilitate interoperability
 Stimulate the adoption of good practices
 Bolster the involvement of players still hesitant about open-data
initiatives, crowdsourcing, data mashups and the new era of
geospatial data ubiquity
 30

The concepts underlying geospatial data quality

Section Overview

 The inherent uncertainty of geospatial data

 The perspectives of geospatial data quality:
 Internal quality
 External quality
 Perceived quality
 Metaquality
 The subject-matters of geospatial data quality
 The management of risks associated to geospatial data
quality
 The dissemination of information on geospatial data quality
and risks of usage
 The standards supporting geospatial data quality and risk
management
 31

The inherent uncertainty of geospatial data

 The process of producing a geospatial dataset induces

uncertainty:
 Loss of details
 Goal dependency
 Model-maker dependency
 Context dependency
 Translations between cognitive and physical models are not
straightforward
 Modeling and communication rules are rarely unequivocal
 32

The inherent uncertainty of geospatial data

 Models supporting geospatial datasets are only

approximations of the real world:
 Conceptual uncertainty
 Is this object a “road” or a “path”? Is this a “path” or not?
 Descriptive uncertainty
 Is this building quality “Standard” or “Standard +”
 Location uncertainty (in space and time)
 ± 5 meters, ± 1day
 Meta-uncertainty
 95% confidence error ellipses

 These four orders of uncertainty combine to generate the

total uncertainty of the dataset
 33

The perspectives of geospatial data quality

Internal quality (ISO 19157): internal characteristics of a
geospatial dataset (i.e. the intrinsic properties resulting
from data production methods)
 Completeness: presence or absence of features, their attributes
and relationships
 Omission or commission

 Logical consistency: degree of adherence to logical rules of data

structure, attribution and relationships
 Conceptual consistency, domain consistency, format consistency,
topological consistency

 Positional accuracy: accuracy of the position of features within a

spatial reference system
 Absolute accuracy, relative accuracy, gridded data position accuracy
 34

The perspectives of geospatial data quality

Internal quality (ISO 19157)

 Thematic accuracy: quality of the thematic attributes and of the

classifications of features and their relationships
 Classification correctness, non quantitative attribute correctness,
quantitative attribute accuracy

 Temporal quality: quality of the temporal attributes and temporal

relationships of features
 Accuracy of time measurement, temporal consistency, temporal
validity
 35

The perspectives of geospatial data quality

External quality (ISO 19157)

 Fitness for use: degree of agreement between data

characteristics (i.e., internal quality) and the explicit and/or
implicit needs of a user for a given application in a given
context

 Usability: based on user’s requirements, all internal quality

elements of ISO 19157 may be used to evaluate usability
 36

The perspectives of geospatial data quality

Perceived quality

 Within a consumer-centered (i.e. B2C and C2C), users may

have a different view of the external data quality of a
geospatial dataset

 Users can rate the dataset based on their perception (using

a 5-star system for example) and write comments

 The global perceived quality is the result of the aggregation

of each individual user perception (bottom-up approach)
 37

The perspectives of geospatial data quality

Metaquality (ISO 19157)

 Information describing the quality of data quality

information:
 Confidence
 Representativity
 Homogeneity

 Metaquality helps estimating the risk related to geospatial

data uncertainties
 38

The subject matters of geospatial data quality

Dataset series level

Dataset level

Subset level

Feature type level

(Adapted from
Devillers, 2004)
Feature instance level
Feature attribute level
Attribute value level
 39

The subject matters of geospatial data quality

 Geospatial data quality can be regarded at various granularity levels
(ISO 19157 data quality scopes)
 Dataset series level (e.g., the National Topographic System of Canada
(NTS))
 Dataset level (e.g., a specific map of the NTS)
 Subset level (e.g., the subset of features included in the North-West
zone)
 Feature type level (e.g., the set of “roads segments” of a topographic
map)
 Feature instance level (e.g., road 138 on a specific topographic map)
 Feature attribute level (e.g., the “Functional road class” of a road
segment)
 Attribute value level (e.g., code value “1” (“Freeway”) for a specific
road segment)
 ISO 19157 provides methods for aggregating quality information from
a single data of a single feature up to the complete dataset
 40

The management of risks associated to

geospatial data quality
 Risk is about the effect of uncertainty:
 Uncertain geospatial data
In the
 Uncertain geospatial data quality
geospatial
 Uncertain geospatial data usages context
 Uncertain expertise of users of geospatial data

 A risk management process serves as a guide on how to

avoid or manage the impacts of uncertainty

 From a legal perspective, using a risk management

approach is necessary to protect both the geospatial data
producer and user
 41

The management of risks associated to

geospatial data quality
 Perfect quality does not exist
 Overall quality involves internal + external + perceived +
metaquality

 Zero risk does not exist

 Risk can be reduced, rarely eliminated

 Quality and risk management are interrelated

 Typically, the higher the quality, the lower the risk to manage

Overall Risk to
Quality manage
 42

The standards supporting geospatial data

quality and risk management
Specific standards
 ISO 19115 Geographic information – Metadata: defines the schema
required for describing geographic information and services by means
of metadata, including geospatial data quality

 The North American Profile of ISO 19115: makes certain optional

fields of ISO 19115 mandatory, supports multiple languages, code
lists, …
 Included in the Treasury Board of Canada Secretariat Standard on Geospatial Data
 Established as a National standard by the Canadian General Standards Board

 ISO 19131 Geographic information – Data product specification:

specifies requirements for the definition of geographic data products,
or user requirements, based upon the concepts of other ISO 19100
International Standards, such as data quality
 Producers: product specifications
 Users: product requirements
 43

The standards supporting geospatial data

quality and risk management
Specific standards (cont.)
 ISO 19157 Geographic information – Data quality: establishes
principles for describing the quality of geographic data by:
 Defining components for describing data quality
 Specifying components and content structure of a register for data quality
measures
 Describing general procedures for evaluating the quality of geographic
data
 Establishing principles for reporting data quality

 ISO/TS 19158 Geographic information – Quality assurance of data

supply: provides a quality assurance framework for the producer and
customer in their production relationship
 44

The standards supporting geospatial data

quality and risk management
Generic standards
 ISO 9000 – Quality management: family of standards that provide
guidance and tools for organizations that want to ensure that their
products and services consistently meet customers’ requirements and
that quality is consistently improved

 ISO 31000 Risk management – Principles and guidelines: provides

principles, framework and a process for managing risk
 IEC 31010 Risk management – Risk assessment techniques
 ISO Guide 73 Risk Management - Vocabulary
 45

The dissemination of information on

geospatial data quality and risks of usage
 The traditional way to communicate quality information is the use of
metadata
 Designed by experts for experts (B2B context), metadata are less appropriate for
other types of actors, particularly for the general public (B2C and C2C mass
markets)
 The process of communicating data quality and risks must be
adjusted for all audiences with new vocabularies, methods and
documentation products
 Since quality and risk will have different values in different usages, the information
will differ for each user/usage

-> Writing proper documentation helps producers to meet their legal

duty for information, advice, and warnings
-> “Good data documentation and well drafted disclaimers and
agreements will minimize data misuse and abuse” (National States
Geographic Information Council, 2011)
 46

Geospatial data quality and risk

management: the fundamentals
Section Overview

 Geospatial data quality management

 Risk management for inappropriate usage of geospatial data
 Communication about geospatial data quality and risks of
usage
 47

Geospatial data quality management concepts

 Geospatial data quality management is the activity of:
 Defining the required quality of needed data
 Defining, implementing and controlling the necessary
steps to ensure quality criteria are met
 Evaluating, documenting and disseminating quality
information
 48

Geospatial data quality management

process
DEFINING THE
REQUIRED QUALITY

STEPS TO ENSURE
DISSEMINATING QUALITY
QUALITY - Definition DOCUMENTING
INFORMATION - Implementation
- Control

EVALUATING
 49

Geospatial data quality management process

in context
The Complete Cycle of Geospatial Data Quality
for a Spatially-Enabled Society Users
Producers
DESCRIBING CONTEXT

DEFINING NEEDS/REQUIREMENTS
(ISO 19131)
(ISO TC 145/ISO 3864-2)
PROPERLY WITH ALL

Selecting/Producing a dataset
COMMUNICATING

REVIEWING AS
MONITORING
NECESSARY
PLAYERS

EVALUATING DATA QUALITY

(ISO 19157)

dataset
accepted?
No
Yes
MANAGING RISK of
inappropriate use of geospatial
data (ISO 31000)
 50

Geospatial data quality management concepts

From a producer point of view (B2B or B2C contexts)

 Geospatial data quality and risks of inappropriate use must

be managed at each phase of a data product life-cycle
(production or update process):
Quality Management

Design Implementation Production Delivery Usage

Risk Management
 51

Geospatial data quality management

From a producer point of view (B2B or B2C contexts)
 Geospatial data quality must be managed at each phase of a data product life-
cycle:

New/Update Design

OR

Application schema
(ISO 19109)
Schema for coverage
geometry and functions (…)
Feature catalog (ISO 19123)
(ISO 19110)

Product specifications (ISO

19131, referring to
appropriate other ISO 19100
series standards)
 52

Geospatial data quality management

From a producer point of view (B2B or B2C contexts)
 Geospatial data quality must be managed at each phase of a data product life-
cycle:

New/Update Implementation Production

Quality
evaluation/control
(ISO 19157)

Integrity constraints (…)

Metadata (ISO 19115,

including quality
information)
 53

Geospatial data quality management

From a producer point of view (B2B or B2C contexts)
 Geospatial data quality must be managed at each phase of a data product life-
cycle:

New/Update Delivery Usage

Product documentation:
User manual,
(…) Well-documented data or
Warnings, and other
quality-aware application
user-centered
documentation
 54

Geospatial data quality evaluation

 Geospatial data quality evaluation can be defined as a

process used to determine whether a geospatial data
product meets the objectives with regards to:
 Product specifications (Producers, internal quality)
 Product requirements for a planned use (Users, “usability”,
external quality or fitness-for-use)

 Can be:
Formal Semi-formal Informal
(ISO 19157) (text, 5-star
rating, etc.)

Experts Non-
experts
 55

Geospatial data quality evaluation

B2B, experts (some B2B cases, B2C, C2C)
Specify (ISO 19157)

Against
Data quality units Informal (inspired by ISO 19157)
specifications/ Against
requirements needs
Data quality measures Informal quality evaluation (inspired
(ISO 19131,
used to by ISO
formally Data quality evaluation 19131)
describe the procedures
needs)
Evaluate (ISO 19157)

Output of quality evaluation Output of metaquality

evaluation

Report (ISO 19157)

Report quality as metadata

Opt. data quality report

 56

Risk management in context

The Complete Cycle of Geospatial Data Quality
for a Spatially-Enabled Society Users
Producers
DESCRIBING CONTEXT

DEFINING NEEDS/REQUIREMENTS
(ISO 19131)
(ISO TC 145/ISO 3864-2)
PROPERLY WITH ALL

Selecting/Producing a dataset
COMMUNICATING

REVIEWING AS
MONITORING
NECESSARY
PLAYERS

EVALUATING DATA QUALITY

(ISO 19157)

dataset
accepted?
No
Yes
MANAGING RISK of
inappropriate use of geospatial
data (ISO 31000)
 57

Risk management concepts

 Risk management is the activity of directing and
controlling what an organization does to minimize
unexpected impacts on its objectives
 Key principle: zero risk does not exist !
 Risk management implies balancing the efforts to
prevent unexpected outcomes with potential negative
impacts of such outcomes for stakeholders (including
consumers)

Efforts Unexpected
negative
impacts
 58

Risk management process

(based on ISO 31000)
DESCRIBING
CONTEXT

ASSESSING RISK
COMMUNICATING MONITORING +
- Identification
PROPERLY TO ALL REVIEWING AS
- Analysis
PLAYERS NECESSARY
- Evaluation

BUILDING RISK
RESPONSES
- Mitigation
- Avoidance
- Transfer/sharing
- Acceptation
 59

Risk management concepts

From a producer point of view (B2B or B2C contexts)

 Risks of inappropriate use and geospatial data quality must

be managed at each phase of a data product life-cycle
(production or update process):
Quality Management

Design Implementation Production Delivery Usage

Risk Management
 60

Risk management concepts

ISO 31000 (2009)
 Generic concepts and vocabulary about risk management
 Approach to stimulate enterprises to develop a risk management
culture and to ask the right questions at the right time (not “after”
harm)
 NOT a guide about good practices since they vary among domains
 For all types of projects
 Small, large
 Individual, company, government, …
 Proposes a new definition of RISK more in sync with business
practices and court decisions regarding liability
 Adds 3 tasks to previous version of standard:
 Context establishment
 Communication and consultation
 Monitoring and review
 NOT for certification or licensing purposes
 61

Concepts: ISO 31000 vocabulary comes from ISO

Guide 73:2009 Risk Management - Vocabulary

 Risk: “effect of uncertainty on objectives”

 Effect: deviation from the expected (positive and/or negative)
 Uncertainty: deficiency of information
 Objectives: economic, environment, health, …
 Risk is often expressed as a combination of the positive or negative
consequences of an event and their likelihood of occurrence
 E.g., most surfaces of pesticide spray are overestimated when using
cadastral parcels area, resulting potentially in a large proportion of
undervalued concentration of pesticide and larger number of farmers
meeting standards than should be. The risk of faulty decisions is high in
most regions because of topography and land use. Since it is a systematic
error, not a random one, there is a high risk that our environmental
objectives be affected negatively (example adapted from Edoh-Alove
et al, 2015)
 62

Risk management: Establishing the context

 General context:
 Objectives: can be economic, environmental, public health, security, …
 Scope: can be project, organisation, community, ...
 External context:
 Local, regional, national, international
 Legal and regulatory requirements
 Stakeholders’ perceptions
 Micro and macro economy
 Social and political environment
 Competition, trends, ...
 Internal context:
 Organisational culture, governance, standards, structure and strategy
 Commitments, contractual relationships
 SWOT analysis (Strengths, Weaknesses, Opportunities, Threats)
 Specific context of the targeted scope:
 Risk management objectives for targeted scope
 Resources, time required, project management
 Depth, breath, inclusions, exclusions, responsibilities
 Methodologies
 Risk criteria, measures, tolerance levels, decisions to make
 63

Risk management: Identifying risks

 Build a comprehensive list of risks that might impact
(good or bad) the achievement of objectives
 I.e. reasons why objectives could potentially not be reached
 Find and describe:
 Sources of risk (under control or not)
 Economic, social, political, natural, markets, technological,
operational, human, legal, …
 Known, unknown, emerging
 Impacts and cumulative effects
 Possible scenarios
 People with appropriate knowledge should be involved
(e.g., experts, users, support service specialists,
consultants, …)
N.B. the complementary IEC 31010 standard provides guidance on risk assessment
techniques
 64

Risk management: Analysing risks

 Comprehend the nature of identified risks by determining
their causes, sources, consequences on objectives, likelihood
to happen and interdependence
 Based on historical data and/or extrapolation and/or prediction
 Consider existing controls
 Consequences can be tangible or intangible
 Can be qualitative or quantitative
 Can be undertaken at various levels of detail
 Comprehend the level of identified risks by combining their
likelihood to happen and consequences
 Indicate the level of confidence in this determination

N.B. the complementary IEC 31010 standard provides guidance on risk assessment
techniques
 65

Risk management: Evaluating risks

 Comparison between the Level of risk obtained during Risk
Analysis and the Risk criteria established when defining the
Context
 Money
 Human casualties
 Delays
 Environmental disaster
 Etc.
 This comparison helps decision-makers to select strategies for
risk treatments and their prioritization
 Risk cannot be tolerated, treatment is essential
 Risk can be tolerated, needs to be monitored
 Risk is negligible, to be observed
 Must consider legal, regulatory and other requirements

N.B. the complementary IEC 31010 standard provides guidance on risk assessment
techniques
 66

Risk management: Treating risks

 Risk treatment implies selecting and implementing one or a
combination of strategies to modify a risk in order to reach
accepted levels of tolerance
 4 categories of strategies can be used:
 Mitigation: actions to eliminate or reduce consequences or their
likelihood of occurring
 Avoidance: eliminate activity to eliminate risk
 Transfer/sharing: shift impact to another entity in part or entirely
 Acceptance: voluntarily accept and take risk
N.B. ignoring a risk = informally accepting a risk
 Several alternatives exist to treat any given risk, they differ in:
 Costs
 Delays
 Efficiency
 Must balance efforts vs. benefits for all stakeholders with regards
to the objectives set forth in the Context
N.B. the complementary IEC 31010 standard provides guidance on risk assessment
techniques
 67

Risk Management: Communicating & consulting

 With internal and external stakeholders

 Address issues about the contexts and interest of stakeholders,
the causes and origins of identified risks, their good and bad
consequences, their level, risk criteria and levels of tolerance,
the treatments that already exist or that will be implemented,
their monitoring and review, etc.
 Examples:
 Joint Committees with various expertises
 Web-based forum
 Paper/on-line user manual in a language understandable by target
readers (Gervais, 2004)
 Training
 FAQ and info@quality.yourorganisation
 Etc.
 68

Risk management: Monitoring & review

 Regular surveillance of the risks and the success of their

treatment
 Detect changes in contexts
 Detect emerging risks
 Seek continuous improvement
 69

Communication about geospatial data

quality and risks of usage
 In a B2B (experts) context:
 Communication products and services that were identified in the B2B contract
 Data quality metadata (ISO 19115, ISO 19157)
 Initial specifications as additional information for new external expert users
 Depending upon the detailed context: data quality report (ISO 19157) potentially for each type
of usage (to complement the metadata)
 Optional: user manual and other communication products and services offered in B2C and C2C
contexts
 In B2C and C2C contexts:
 User manual highly recommended (Gervais, 2004), it may include:
 Warnings (including symbols (ISO TC 145/ISO 3864-2))
 Disclaimer
 Recommended and non-recommended usages
 and much more
 License
 Guarantee
 …
 A combination of communication methods as described in the Risk
Treatment strategies in a later section of this presentation
 70

Putting theory into practice

Section Overview

 Managing geospatial data quality in practice

 Managing the risks of geospatial data usage in practice
 Communicating about geospatial data quality and the related
risks of usage in practice
 71

Managing geospatial data quality in practice

ISO 19109:2005 Geographic information – Rules
for application schema
• Conceptual modeling of features and their properties from a
universe of discourse
• Definition of application schemas
• Use of the conceptual schema language for application schemas
• Transition from the concepts in the conceptual model to the data
types in the application schema
• Integration of standardized schemas from other ISO geographic
information standards with the application schema
ISO 19157: verify the
DQ_LogicalConsistency of the
schema

Example:
Road network model (extract from
Levesque, Bédard, Gervais, & Devillers,
2007)
Warning symbols to
highlight potential
problems
 72

Managing geospatial data quality in practice

ISO 19110:2005 Geographic information –
Methodology for feature cataloguing
• Methodology for cataloguing feature types
• Specifies how the classification of feature types is organized into a
feature catalogue and presented to the users of a set of
geographic data.

ISO 19157: verify the

DQ_LogicalConsistency
of the catalog

Example:
The CanVec+ Feature
Catalog (extract)

Add a section to
describe warnings
http://ftp2.cits.rncan.gc.ca/pub/canvec/doc/CanVec_feature_catalogue_en.pdf
 73

Managing geospatial data quality in practice

ISO 19123:2005 Geographic information – Schema
for coverage geometry and functions
• Conceptual schema for the spatial characteristics of coverages
• Relationship between the domain of a coverage and an associated
attribute range

D igital OrthoIm agery

ISO 19157: verify the 1..*

CV_Continuous GridCoverage

DQ_LogicalConsistency of
AreaOfInteres t Image + interpolationType : CV_InterpolationMethod = neares tNei ghbor
+ domainExtent : EX_Extent
1 + rangeType : RecordType

the schema +collection

0..1
+evaluator

Example: +element

Top level classes for digital

1..*

Band Pixel

orthoimagery (from Maitra, 1..*

2004)
+source
1
CV_GridValues Matrix
(f ro m Ele v ation )

1 + gridRange : CV_GridRange
DNValue + val ues : Sequence<Record>
+ sequencingRule : CV_SequenceRule
+ startSequence : CV_GridCoordinate

73
 74

Managing geospatial data quality in practice

ISO 19131:2007 Geographic information – Data
product specifications (with ISO 19131:2007/Amd 1:2011
Requirements relating to the inclusion of an application schema and feature
catalogue and the treatment of coverages in an application schema)
• Requirements for the specification of geographic data products, based
upon the concepts of other ISO 19100 International Standards
• References application schema, feature catalog or schema for coverage
geometry

Section in the
ISO 19157: verify the specifications to
DQ_LogicalConsistency describe the expected
quality
of the schema

ISO 19157: express the

expected values for all
DQ_Elements

Example:
The CanVec+ Data Product
Specifications (extract)
http://ftp2.cits.rncan.gc.ca/pub/canvec+/doc/CanVec+_product_specifications.pdf
74
 75

Managing geospatial data quality in practice

Integrity constraints (spatial, temporal, thematic)
• Intra-field (e.g., values of a numeric field must be between 0 and 1)
• Inter-fields (e.g., if the value of the road classification attribute is
“national”, then the value of the maximum speed attribute cannot be
null)
• Intra-feature (e.g., the date of an updated house assessment cannot
be lower than the date of the older house assessment)
• Inter-feature (e.g., the size of a “building” must be smaller than the
size of the “parcel” it is built on)
• Intra-feature class (e.g., “building” cannot intersect “building”)
• Inter-feature classes (e.g., “road” cannot cross “lake”)
ISO 19157: Integrity • Intra-theme (e.g., “river” can connect “canal”)

constraints may help in • Inter-theme (e.g., “dam” can share geometry with “road”)

controlling all
DQ_Elements

Example:
Constraint repository
(adapted from Normand,
1999)

75
 76

Managing geospatial data quality in practice

ISO 19157:2013 Geographic information – Data
Quality Examples:

Unit 1: Dataset N, completeness

of hydrants

Measure 1: Number of excess

items

• Components for Full inspection

describing data quality
• Components and content
structure of a register for 0 excess item; pass
data quality measures
• General procedures for 100% confidence
evaluating the quality of
geographic data
• Principles for reporting
data quality DQ_CompletenessCommission

Detailed steps

76
 77

Managing geospatial data quality in practice

ISO 19115-1:2014 Geographic Information – Metadata – Part 1:
Fundamentals
ISO 19115-2: 2009 Geographic Information – Metadata – Part 2: Extensions
for imagery and gridded data
North American Profile of ISO 19115:2003 — Geographic Information —
Metadata (NAP — Metadata)
ISO/TS 19139:2007 Geographic Information – Metadata – XML schema
implementation
• Mandatory and conditional metadata sections, metadata entities, and
metadata elements
• The minimum set of metadata required to serve most metadata
applications (data discovery, determining data fitness for use, data
access, data transfer, and use of digital data and services)
• Optional metadata elements to allow for a more extensive standard
ISO 19157: document description of resources, if required
the values for all • A method for extending metadata to fit specialized needs

DQ_Elements

Example:
The CanVec+ 082C
metadata (extract)

Data quality
elements in metadata
77
 78

Managing geospatial data quality in practice

Proper communication typically combines several
information products/services and is a direct result of the
risk management strategies adopted

Strongly recommended: User manual (with the content

suggested by Gervais (2004) based on legal
considerations):
• License
• Guarantees
• Installation
• Product description
ISO 19157: values for • Resolution (spatial, temporal, descriptive) of the data
DQ_Elements will • General advice
influence the product • Functional specifications

documentation • Recommended uses

• Non-recommended uses
• Warnings and safety
• Troubleshooting
• Technical specifications

ISO 19115 metadata already contains part of this information, but in a technical
jargon usually unintelligible for most users (see Gervais 2004 for the
correspondence)
 79

Managing geospatial data quality in practice

Examples of product documentation:

Example of a Guarantee section of a user manual (extract from Gervais, 2004)

Example of a Warnings and safety section of a user manual (extract from Gervais,
2004)
ISO 19157: values for
DQ_Elements will
influence the product
documentation Example of a Troubleshooting section of a user manual (extract from Gervais, 2004)

Example of a data Example of a

product User Geospatial data
manual (Statistics quality good
Canada, 2009) practices guide
(Statistics Canada,
2009)
 80

Managing geospatial data quality in practice

Warnings: ISO TC 145/ISO 3864-2 Graphical
symbols
• Establishes the principles for preparation,
coordination and application of graphical
symbols
In Geomatics:
• Numerous standards exist for map products (e.g.
topographic maps, nautical charts, geology maps)
• Standards exist for copyright (e.g. Creative Commons)
• However, no standard exists yet for geospatial data
ISO 19157: values for quality analysis and risk management
DQ_Elements will
Symbols and labels are powerful ways to convey the
influence the product
meaning of risk:
documentation
 Type of risk (danger or positive action)
 Level of risk
 Description of risk
 Actions to take in face of consequences
 81

Managing geospatial data quality in practice

ISO 19157: values for http://www.safetysign.com/help/h40/safety-header

DQ_Elements will
influence the product
documentation

Example:
Use of symbols to facilitate
the reading of a quality report
(private report by Gervais,
Bédard and Larrivée, 2007)
 82

Managing geospatial data quality in practice

Continuation of means put into place for a quality-aware
delivery of the data
E.g., quality-aware application
• Displays warnings to users according to the data they consult or
the operations they conduct on the data
• Based on metadata

Example of a context-sensitive
warning of inconsistency after a
query in a quality-aware application
(Gervais et al., 2009). The warning
contains 3 parts recommended by
ISO: level of risk, nature of problem,
action to solve problem
ISO 19157: values for
DQ_Elements will
dictate how the product
should be used

Examples: Example of VQI using a 5-

star rating system (from
Dataset is rated using a (Koistinen, 2015))
number of stars
 83

Geospatial data quality evaluation

B2B, experts (some B2B cases, B2C, C2C)
Specify (ISO 19157)

Against
Data quality units Informal (inspired by ISO 19157)
specifications/ Against
requirements needs
Data quality measures Informal quality evaluation (inspired
(ISO 19131,
used to by ISO
formally Data quality evaluation 19131)
describe the procedures
needs)
Evaluate (ISO 19157)

Output of quality evaluation Output of metaquality

evaluation

Report (ISO 19157)

Report quality as metadata

Opt. data quality report

 84

Evaluating geospatial data quality in practice

ISO 19157:2013 Geographic information – Data
quality – Data quality units

Data quality unit = 1 scope + N data quality elements

• Scope (MD_Scope): specifies the extent, spatial and/or temporal,
and/or common characteristic(s) that identify the data on which data
quality is to be evaluated
• Data quality elements (DQ_Element) from ISO 19157:2013):

Examples:
Quality unit 1:
MD_Scope: dataset
DQ_Elements: DQ_LogicalConsistency,
DQ_Completeness

Quality unit 2:
MD_Scope: feature type (hydrant)
DQ_Element:
DQ_QuantitativeAttributeAccuracy

See 19157:2013 for more examples

 85

Evaluating geospatial data quality in practice

ISO 19157:2013 Geographic information – Data
quality – Data quality measures

A data quality element should refer to one measure

only, by means of a measure reference
(DQ_MeasureReference)
• measureIdentification
• nameOfMeasure
• measureDescription

• List of standard measures provided in ISO 19157

• New measures can be created

Examples (ISO 19157, annex D):

For DQ_CompletenessCommission: For DQ_CompletenessOmission:
Excess item Missing item
Number of excess items Number of missing items
Rate of excess items Rate of missing items
Number of duplicate feature instances

See 19157:2013 for more examples

 86

Evaluating geospatial data quality in practice

ISO 19157:2013 Geographic information – Data
quality – Data quality evaluation procedures

Data evaluation procedure = 1,N data evaluation methods

Data quality evaluation methods (DQ_EvaluationMethod)

can be divided into two main classes: direct and indirect
• Direct evaluation methods compare the data with internal and/or external
reference information
• Indirect evaluation methods infer or estimate data quality using information
on the data such as lineage

• DQ_FullInspection
• DQ_SampleBasedInspection
• DQ_IndirectEvaluation

Example:
Area guided non-random sampling method
X X X
X X X X X X X
X X X X X X X
X X X
See 19157:2013 for more examples
 87

Evaluating geospatial data quality in practice

ISO 19157:2013 Geographic information –
Data quality – Output of quality evaluation

At least one data quality result (DQ_Result)

is provided for each data quality element
• DQ_QuantitativeResult
• DQ_ConformanceResult
• DQ_DescriptiveResult

Examples:
DQ_QuantitativeResult
(DQ_CompletenessCommission), Number of excess items: 3
DQ_ConformanceResult
(DQ_CompletenessCommission), Number of excess items: pass
DQ_DescriptiveResult
(DQ_LogicalConsistency), Conceptual schema compliance: “The rules of the CanVec+ conceptual schema are
all recorded and validated in the source database containing the CanVec+ product. This approach ensures the
conceptual consistency between the conceptual schema and the CanVec+ product.” (from CanVec+ 082C
metadata (extract))

See 19157:2013 for more examples

 88

Evaluating geospatial data quality in practice

ISO 19157:2013 Geographic information –
Data quality – Output of metaquality
evaluation

Metaquality elements are a set of quantitative and

qualitative statements about a quality evaluation and its
result:
• DQ_Confidence: trustworthiness of a data quality result
• DQ_Representativity: degree to which the sample used has
produced a result which is representative of the data within the
data quality scope
• DQ_Homogeneity: expected or tested uniformity of the results
Examples (from ISO 19157): obtained for a data quality evaluation
DQ_Confidence
Standard deviation or a confidence interval on a given confidence level.
DQ_Representativity
All the geographic zones and concerned time periods are covered and the population is
sufficiently large
DQ_Homogeneity
Comparison of the evaluation results of several segments of a global data set expressed using
root mean square errors
See 19157:2013 for more examples
 89

Evaluating geospatial data quality in practice

ISO 19157:2013 Geographic information –
Data quality – Report quality as metadata

Data quality is reported as metadata in compliance with

Clause 7, Clause 10, Annex C, ISO 19115-1:2014 and
ISO 19115-2:2009

Example:

The CanVec+ 082C

metadata (extract)

See 19157:2013 for more examples

 90

Evaluating geospatial data quality in practice

ISO 19157:2013 Geographic information –
Data quality – Data quality report
In order to provide more details or to present the
information in an easier to understand format than
reported as metadata, a standalone quality report may
additionally be created
Usually one report per usage/user
Examples: • The structure of the report is free.
Standalone data quality
report: commission per
feature class (extract from
ISO 19157:2013)

Summarized feature class

quality evaluation (adapted
from a private report by
Gervais, Bédard and Larrivée,
2007)

See 19157:2013 for more examples

90
 91

Evaluating geospatial data quality in practice

Informal data quality evaluation
• For producers: in B2B and B2C contexts, less formal processes
to evaluate and report data quality should be inspired by the
ISO 19157 approach

• For consumers in the B2C and C2C context: various means are
used, unknowingly following ISO 19157 rationale in a less
rigorous manner. They typically fit their measurement method
with their quality representation method.

Examples:
Use of a 5-star rating
system to rate the
representation of
buildings in a virtual
globe environment
(from (Jones, 2011))

Use of a 5-star rating

system in a SDI
environment (from
(Koistinen, 2015))
 92

Risk management process

(based on ISO 31000)
DESCRIBING
CONTEXT

ASSESSING RISK
COMMUNICATING MONITORING +
- Identification
PROPERLY TO ALL REVIEWING AS
- Analysis
PLAYERS NECESSARY
- Evaluation

BUILDING RISK
RESPONSES
- Mitigation
- Avoidance
- Transfer/sharing
- Acceptation
 93

Managing the risks of geospatial data usage

in practice
ISO 31000:2009 Risk Management –
DESCRIBING Describing the risk management context
• General context
CONTEXT • External context
• Internal context
• Specific context of the targeted scope

Understand how the planned activity fits into the

wider organization and market/society and the
organization’s approach to risk management, in
order to scope the risk management strategy
Example:
There are 3 other publicly available sources of data similar to ours. Their cost and overall accuracy is
similar to ours and clients have difficulty to choose the one that best fit their needs. The evolution of
the market is uncertain as well as the actions of the competition. By implementing a new data
quality and risk management strategy, it will help to improve significantly the communication with
potential clients and help them understand how our data can fulfill their geospatial needs. A potential
client who knows better is reassured and has more chances of becoming client. Our objective is to
increase our market share by 10% within 2 years. If we implement the strategy, the costs and risks
are ... If we don't, the risks are...
 94

Managing the risks of geospatial data usage

in practice
ISO 31000:2009 Risk Management –
Identifying risks for the users of data
ASSESSING RISK This step will generate a comprehensive list of risks of
- Identification
Identification inappropriate use of geospatial data
- Analysis Conducted using:
- Evaluation • Analysis of existing documentation (e.g.,
specifications, contracts, and task flow charts)
• Interviews
• Brainstorming sessions
• Collaborative approach with users, …
Example: Feature
List of potential risks / Identified Risks
Attribute
of inappropriate use R-1: the user may think that the whole region delimited by the cadastral boundaries
Cultivated
of geospatial data was cultivated. Some areas may not be cultivated, such as woodland, rocky button or
parcel
related to areas near cadastral boundaries.
R-2: Floodplains are vague data. The user would think that the provided boundaries
features/attributes in Floodplain are accurate whereas they are fuzzy and large boundaries are not represented as such
a geospatial dataset on the map.
R-3: the areas where the pesticide is spread have large and fuzzy boundaries and
(extract from Grira, uncertain location within the plot (because of the techniques and the methods of
2014) Pesticide spread
pesticide spreading). The user could think that the area is accurate whereas positional
accuracy is not considered in the area calculation.
area Note: the uncertainty for R-3 is related to the pesticide spreading zone, i.e. its
boundaries and its location. However, the uncertainty for R-1 is related to the plot (its
boundaries) where the spreading zone is located.
 95

Managing the risks of geospatial data usage

in practice
ISO 31000:2009 Risk Management –
ASSESSING RISK Analyzing risks
Causes and sources of risks
- Identification •
• Consequences
- analysis
Analysis • Likelihood of occurring
- Evaluation Conducted using:
• Analysis of lessons learned from previous projects
• Simulation methods
• Probabilistic analysis, …

Feature
Identified
Example:
/ Impact of Risk Probability of Occurrence
Risks
Attribute
Analysis of Cultivated
Strong overestimation of the
ratio quantity of pesticide / Medium
potential risks parcel
R-1
hectare (high)
of inappropriate Strong underestimation of the
use of Floodplain R-2 quantity of pesticides that might Medium
geospatial data be present in water (high)

(extract from Strong underestimation of the

Grira, 2014) Pesticide
quantity of pesticides that might
spread R-3 Medium
be present in water (high)
area
 96

Managing the risks of geospatial data usage

in practice ISO 31000:2009 Risk Management –
Evaluating risks
ASSESSING RISK • Prioritize risks according to level of tolerance
• Will help to select strategies for risk treatments
- Identification • Must consider legal, regulatory and other requirements
- Analysis Conducted using:
- evaluation
Evaluation • Ranking matrix

Feature Probability
Identified Overall Risk
Example:
/ Impact of Risk of
Risks Evaluation
Attribute Occurrence
Evaluation of Cultivated
Strong overestimation of
the ratio quantity of Medium High
potential risks of parcel
R-1
pesticide / hectare (high)
inappropriate Strong underestimation of
use of geospatial the quantity of pesticides
Floodplain R-2 Medium Medium
data (extract that might be present in
water (high)
from Grira,
Strong underestimation of
2014)
the quantity of pesticides
Pesticide
R-3 that might be present in Medium Medium
spread area
water (high)
 97

Managing the risks of geospatial data usage

in practice
BUILDING RISK ISO 31000:2009 Risk Management –
RESPONSES Mitigating risks
- mitigation Strategies that modify actions to eliminate or reduce
Mitigation •
consequences or their likelihood of occurring
- Avoidance • May have significant impact on costs and additional
- Transfer/sharing efforts
- Acceptation

Examples:
• Improve database design/dataset structure
• Improve the quality control of the dataset (e.g., add integrity constraints)
• Use standards (e.g., for data quality and risk management interoperability)
• Properly inform users in a language they understand (highly recommended)
• Provide a user manual
• Offer a 1-800 help line or Ask@yourcompany.com
• List target usages and non-recommended usages
• Provide a Guide of good practices
• Train users, …
• Conduct tests on the dataset (users)
• Compare with another dataset (users)
• (…)
 98

Managing the risks of geospatial data usage

in practice
BUILDING RISK ISO 31000:2009 Risk Management –
RESPONSES Mitigating risks
- mitigation
Mitigation
- Avoidance Examples:
- Transfer/sharing
- Acceptation Communicating risks of
inappropriate use of geospatial
Communicating risks of data at the usage phase (from
inappropriate use of Gervais et al., 2009)
geospatial data at the
design phase (from
Levesque et al., 2007)

Web-based forum (from Grira, 2014)

 99

Managing the risks of geospatial data usage

in practice
BUILDING RISK ISO 31000:2009 Risk Management –
RESPONSES Avoiding risks
- Mitigation • Strategies that eliminate the activity to eliminate risk
- Avoidance
Avoidance
- Transfer/sharing
- Acceptation

Examples:
• Stop distributing or using the dataset or a part of
• Eliminate a category of users
• Eliminate a data provider
• Explicitly and clearly forbid a given usage
• (…)
 100

Managing the risks of geospatial data usage

in practice
BUILDING RISK ISO 31000:2009 Risk Management –
RESPONSES Transferring and sharing risks
- Mitigation • Strategies that move to, or share a risk with, other
parties in order to reduce it or to eliminate it
- Avoidance
-Transfer/sharing
Transfer
- Acceptation

Examples:
• Buy an insurance
• Obtain the dataset from a broker who can give advice related to its use
• Use a dataset with a guarantee that explains clearly risk sharing (who is responsible of what)
• The content of a guarantee for geospatial products is described in a paper by Plante and
Gervais to be published in Geomatica in 2015.
• Have the dataset evaluated by an expert
• Replace a B2C strategy with a B2B strategy for your business by contracting a data broker who
will offer the B2C strategy
• Have the data quality evaluated by an external expert for the new usages
• (…)
 101

Managing the risks of geospatial data usage

in practice
BUILDING RISK ISO 31000:2009 Risk Management –
RESPONSES Accepting risks
- Mitigation • Strategy that voluntarily accept the risks
- Avoidance
- Transfer/sharing
Acceptation
- Acceptation

Example:
• Use the dataset no matter what the risks are and do nothing about it, i.e. take
the risk
 102

Managing the risks of geospatial data usage

in practice
ISO 31000:2009 Risk Management –
COMMUNICATING Communicating risks
PROPERLY TO ALL • Communicating with the persons in charge of offering
data/services, the persons in charge of the data
PLAYERS production, and the users
• Developing the information products identified in the
risk treatment strategies, especially:
• B2C + C2C: paper or on-line user manuals with
previously recommended content
• B2B: embed risk-related information within data
quality metadata and report
• Using a vocabulary adapted to the various audiences
• Properly informing users: distribute the information
products; keep users informed of new context
elements, new quality controls, new quality
evaluations, new risk management strategies
implemented, new usages, new restrictions, new good
practices, etc.
• Promoting joint committees with various expertises
• Supporting Frequently Asked Questions (FAQ), 1-800
info lines, info@quality.yourorganisation
• Offering training, webinars
• (…)
…
 103

Managing the risks of geospatial data usage

in practice
ISO 31000:2009 Risk Management –
COMMUNICATING Communicating risks
PROPERLY TO ALL
PLAYERS

Examples:
Quality metadata (the
CanVec+ 082C
metadata (extract))

Standalone data quality

report: commission per
feature class (extract
from ISO 19157:2013)
 104

Managing the risks of geospatial data usage

in practice
ISO 31000:2009 Risk Management –
MONITORING +
Monitoring and reviewing risks
REVIEWING AS • To oversee and systematically evaluate, using metrics, the
NECESSARY effectiveness of actions taken
• To update the initial list of identified risks and their
characteristics
• To gather information useful for the development or the update
of risk response strategies
• To review some aspects related to the planning process of risk
management as a whole
• To collect feedback about quality (e.g., via web-based forum or
5-star VQI Volunteered Quality Information)
• To implement new restrictions, add integrity constraints,
implement new training, build/update a register for quality and
Example: risk management, …

Risk register (adapted from Grira, 2014)

Feature
Identified Probability of Overall Risk Action
/ Impact of Risk Owner Response Status
Risks Occurrence Evaluation items
Attribute
Strong overestimation of the Follow-
Cultivated Project Mitigate (date)
R-1 ratio quantity of pesticide / Medium High up with
parcel manager (…) open
hectare (high) users
Strong underestimation of the Follow-
Project Mitigate (date)
Floodplain R-2 quantity of pesticides that might Medium Medium up with
manager (…) open
be present in water (high) users
 105

Recommendations
Section Overview

 Recommendations in a Business-to-Business context

 Recommendations in a Business-to-Consumer context
 Recommendations in a Consumer-to-Consumer context
 106

Recommendations in a B2B context

 Challenge: more formal data quality evaluation and reporting
 Recommendations:
 Facilitate communication between experts about data quality by adopting a
common language
 ISO 19157
 ISO 19115-1 (including quality metadata)
 Synthesized quality reports, aggregated quality information, automated Q&A advisory system
 Foster more efficient geospatial data reuse and interoperability by adopting a
common frame of reference regarding quality (i.e. set of concepts)
 ISO 19157
 ISO 19158
 Facilitate contractual agreements by adopting this common frame of reference
and common language
 Decrease the risks of inappropriate use of geospatial data using risk
management concepts and strategies
 ISO 31000
 Encourage high data quality with strong quality assurance procedures and quality
controls (i.e. improve metaquality)
 ISO 9000
 ISO 19157
 ISO 19158
 Quality auditing, certification, user accreditation
 107

Recommendations in a B2B context

 Recommendations (cont.):
 Develop a Guide of Good Practices to visually represent geospatial data quality
(quality maps, quality radars, quality tables, quality warning symbols, etc.)
 Develop a Guide of Good Practices to mitigate the risks of geospatial data use
 Further clarify the roles and responsibilities between contracting parties by
including quality guarantees
 Gradually implement good practices with the help of geospatial data quality
experts
 Develop a series of quality-related products and services such as quality
guarantee, quality certificate, quality audit, quality control and quality assurance
mechanisms, accreditation of quality experts
 108

Recommendations in a B2C context

 Challenge: managing risks for a better protection of consumers (and
providers)
 Recommendations:
 Facilitate geospatial data selection based on users’ needs (cf. external quality) by
providing:
 Lists of recommended and non-recommended uses
 1-800 free line or Ask@yourcompany.com
 Contribute to the advancement of the geospatial community and spatially-
enabled society by offering:
 User manuals written in a language understandable by the target users
 See slide 78 for suggested content
 Emphasize on clear advices and warnings (use symbols, cf. ISO TC 145/ISO
3864-2)
 Real guarantees (as for any other product or service in a mature market)
 Guides of Good Practices
 Synthesized quality reports and aggregated quality information
 Stimulate quality analysis and users’ awareness with:
 Web-based participatory VQI (Volunteered Quality Information) (e.g., 5-stars +
comments)
 Web-based users forums
 Web-based or in-person training
 109

Recommendations in a B2C context

 Recommendations (cont.):
 Reduce the uncertainty related to certain law-related topics by investing into
studies to:
 Understand the new trends and rights regarding privacy, data ownership, copyright,
data vs. service
 Further develop the concept of guarantee (see Plante and Gervais, 2015)
 Further clarify the responsibility of non-experts VGI data contributors
 Clarify responsibilities when geospatial services and data cross borders
 Stimulate legal interoperability of geospatial data (see Uhlir, 2013)
 Rapidly make the move towards becoming a mature mass-market
 Specialized training, innovation, collaboration
 Improve metadata (i.e., easier to use, new types of quality-centered and risk-
centered metadata written for the end-users)
 Increase geospatial data providers’ and users’ awareness of potential risks of
inappropriate usage of geospatial data by gathering, on a crowdsourced web
site, examples of damages
 110

Recommendations in a C2C context

 Challenge: increasing awareness
 Recommendations:
 Rapidly inform developers of public-oriented web-based systems and
smartphone apps about their duty and potential liability with regards to
geospatial data quality
 Facilitate geospatial data selection based on users’ needs (cf. external quality) by
providing:
 Lists of recommended and non-recommended uses
 1-800 free line or Ask@yourcompany.com
 User manuals written in a language understandable by the target users
 See slide 78 for suggested content
 Emphasize on clear advices and warnings (use symbols, cf. ISO TC 145/ISO
3864-2)
 Guides of Good Practices
 Training and collaboration for the providing C (e.g., when publishing data mashups)
 Reduce the uncertainty related to certain law-related topics by investing into
studies to:
 Understand the new trends and rights regarding privacy, data ownership, copyright
 Further clarify the responsibility of data contributors, integrators and distributors
 Clarify responsibilities when geospatial services and data cross borders
 111

Conclusions
 As geospatial data is increasingly being produced and (re)used by
new types of actors, the question of geospatial data quality is
becoming a major concern
 The objective of this guide was to support the Canadian geospatial
community into its efforts to make the spatially-enabled society more
aware of geospatial data quality
 Based on international standards such as ISO 19157 (Geospatial data
quality) and ISO 31000 (Risk management), this guide presented:
 The concepts underlying geospatial data quality
 The management of geospatial data quality
 The geospatial data quality evaluation process in details (based on
ISO 19157)
 The management of risks of inappropriate use of geospatial data
(based on ISO 31000)
 Detailed examples of quality evaluation and risk management
tasks to be undertaken in the B2B, B2C and C2C contexts
Question and Answer Session
How to find the CGDI
Resource Centre
 114

http://www.nrcan.gc.ca/earth-sciences/geomatics/canadas-
spatial-data-infrastructure/8906
 115

Thank you!
Mr. Eric Wright
Geomatics Engineer, CCEO/GeoConnections
Eric.Wright@NRCan.gc.ca

Dr. Yvan Bédard

Senior Scientific and Strategic Advisor, Intelli3 Inc.
Yvan.Bedard@Intelli3.com

To access copies of resources discussed today, please visit:

geoconnections.nrcan.gc.ca

QUESTIONS?

View publication stats