Consensus-building among various stakeholders from different fields is an important issue in order to facilitate policy and decision-making. For consensus-building stakeholders have to know what others are thinking about each other because differences of their viewpoints cause some conflicts. In this paper we propose a consensus-building support system based on ontology exploration. The key ideas consists two steps 1) developing an ontology to provide a base knowledge to be shared among the users (stakeholders), 2) each user explore the ontology according to his/her viewpoint and generate conceptual maps as the result of the exploration, and 3) they know differences of viewpoints through comparison of generated maps. This paper shows an overview of this tool, and discusses its usability and effectiveness through evaluation experiments by domain experts.
Snow Chain-Integrated Tire for a Safe Drive on Winter Roads
A Consensus-Building Support System based on Ontology Exploration
1. IESD2012
9th Oct. 2012, Galway, Ireland
A Consensus-Building Support
System based on Ontology
Exploration
Kouji Kozaki1, Osamu Saito2 and Riichiro Mizoguchi3
1The Institute of Scientific and Industrial Research, Osaka University, Japan
2United Nations University, Institute for Sustainability and Peace
3Research Center for Service Science, Japan Advanced Institute of Science and
Technology
9 Oct 2012 IESD2012 1
2. Outline
Motivation
A Consensus-Building Support System based on
Ontology Exploration
Divergent exploration of an ontology
Consensus-Building Support System based on Ontology Exploration
Experiment for evaluation in biofuel domain
Demo
Concluding remarks
9 Oct 2012 IESD2012 2
3. Motivation
Consensus-building among various stakeholders
It is one of key issues to solve for facilitating their collaboration.
In order to build consensus, it is important to know what others
are thinking about each other because differences of their
viewpoints cause some conflicts.
However, it is difficult to understand different views in particular
when they come from different fields.
Our Approach
We propose an ontology based system which shows
differences of viewpoints by different stakeholders in order to
facilitate consensus-building among them.
This presentation
Consensus-building support system based on ontology
exploration.
Evaluation experiments by domain experts in sustainable science
(environmental) domain (in particular biofuel).
9 Oct 2012 IESD2012 3
4. Our approach: Consensus-Building
Support based on Ontology
Exploration
1) An ontology provides 2) They explore the
Understanding from the a base knowledge to ontology according to
their own viewpoints
be shared among the their viewpoint and
users (stakeholders). generate conceptual
Stakeholder 1 Target World maps as the result.
×
Ontology developer Conceptual
Stakeholder 2 map
Ontology
× Stakeholder 1
Stakeholder 3 ✓
consensus-building
is difficult
Stakeholder 2 Stakeholder 3
✓
It can facilitate consensus- 3) They can understand
building among differences of viewpoints through
stakeholders. comparison of generated maps.
9 Oct 2012 IESD2012 4
5. (Divergent)
Ontology exploration tool
1) Exploration of multi-perspective conceptual chains
2) Visualizations of conceptual chains
Visualizations as
Exploration of an ontology conceptual maps from
different view points
“Hozo” – Ontology Editor
Multi-perspective conceptual chains
represent the explorer’s understanding of
ontology from the specific viewpoint. Conceptual maps
9 Oct 2012 IESD2012 5
6. Node represents Is-a (sub-class-of)
a concept relationshp Referring to
(=rdfs:Class) another concept
slot represents
a relationship
(=rdf:Property)
9 Oct 2012 IESD2012 6
8. Option settings for
exploration
Selected relationships
Kinds of aspects
are traced and shown as
links in conceptual map
property
names
constriction
tracing classes
Conceptual map visualizer Aspect dialog
9 Oct 2012 IESD2012 8
9. Functions for ontology
exploration
Exploration using the aspect dialog:
Manual exploration
Divergent exploration from one concept using the aspect
dialog for each step
Machine exploration
Search path:
Exploration of paths from stating point and ending points.
The tool allows users to post-hoc editing for extracting
only interesting portions of the map.
Change view:
The tool has a function to highlight specified paths of
conceptual chains on the generated map according to given
viewpoints.
Comparison of maps:
The system can compare generated maps and show the
common conceptual chains both of the maps.
9 Oct 2012 IESD2012 9
10. Search
Path Ending point (1)
Selecting of ending points
Finding all possible
paths from stating
point to ending points
Starting point
Ending point (2)
Ending point (3)
9 Oct 2012 IESD2012 10
12. Functions for ontology
exploration
Exploration using the aspect dialog:
Divergent exploration from one concept using the aspect
dialog for each step
Search path:
Exploration of paths from stating point and ending points.
The tool allows users to post-hoc editing for extracting
only interesting portions of the map.
Change view:
The tool has a function to highlight specified paths of
conceptual chains on the generated map according to given
viewpoints. →Differences between
Comparison of maps: Viewpoints of stakeholders
The system can compare generated maps and show the
common conceptual chains both of the maps.
9 Oct 2012 IESD2012 12
13. Consensus-building support based
on ontology exploration
・Display multiple concept
Map
maps
2 ・Highlight common concepts
Map ・Highlight different concepts
1
Map
4
Touch-Table
Map
3
2nd Step: Collaborative workshop
1st Step: Individual concept map creation
9 Oct 2012 IESD2012 13
14. Comparison of conceptual
maps
The system facilitates discussion among stakeholders through
comparison of conceptual maps they generated.
The system integrates conceptual maps generated by the stakeholders
into an integrated map which consists of all paths appeared in the
maps.
In the generated map,
each path is shown in different
color according to stakeholders.
When the same nodes appeared in both of maps
by different stakeholders, they are shown in
graduations of colors corresponds to them.
14
15. Experiments for
Evaluation
Target domain and topics
Biofuel production in sustainability science (environmental
domain) .
An experiment for evaluating ontology exploration
tool by domain experts [Kozaki 2011]
Subjects: 4 domain experts
Goal: To evaluate whether the tool can generate maps which are
meaningful for domain experts.
An experiment of consensus building by role-play
discussion
9 Oct 2012 IESD2012 15
16. Experiment for evaluating
ontology exploration tool
Experimental method
1) The four experts to generated
conceptual maps with the tool in
accordance with condition settings of
given tasks.
2) They remove paths that were
apparently inappropriate from the
paths of conceptual chains included in
the generated maps.
The subjects: 3) They select paths according to their
4 experts in different fields. interests and enter a four-level general
A: Agricultural economics evaluation with free comments.
B: Social science
(stakeholder analysis) A: Interesting
C: Risk analysis B: Important but ordinary
D: Metropolitan environmental
planning
C: Neither good or poor
D: Obviously wrong
9 Oct 2012 IESD2012 16
17. Experimental results
Table.2 Experimental results . l
Number of maps
Number of Path distribution based on general evaluation
generated: 13 selected paths A B C D a
Expert A 2 2
Number of paths evaluated:1 61
Expert A
1
(second time)
A: Expert B
Interesting 307 (49%) 4 1 85%
2
B: Expert B
Important but6 ordinary 22 (36%)
Task 1
3 3
C: Expert C good or poor 8(13%)5
Neither
(second time)
8 1 2
D: Expert D
Obviously wrong 1(2%)
3 1 1 1
Expert A 1 1 E
We can conclude that the tool could generate
Task 2
Expert B 6 1 5 n
Expert C 7 4 1 2 in
maps or paths sufficiently meaningful for experts.
Expert D 5 1 13
c
Expert B 8 4 2 2
n
Number of paths
Task 3 Expert C 4 2 2
Expert D 3 3 p
evaluated: 61
Total 61 30 22 8 1
9 Oct 2012 IESD2012 17
18. Evaluation experiment
Target domain and topics
Biofuel production in sustainability science (environmental
domain) .
An experiment for evaluating ontology exploration
tool by domain experts [Kozaki 2011]
Subjects: 4 domain experts
Goal: To evaluate whether the tool can generate maps which are
meaningful for domain experts.
An experiment of consensus building by role-play
discussion
Subjects: 4 students and 5 domain experts
Goal: To evaluate whether ontology explorations and generated
maps could facilitate a better mutual understanding for
consensus-building among stakeholders.
9 Oct 2012 IESD2012 18
19. An experiment of consensus
building by role-play
discussion
Subjects
Group A: 4 students in environmental engineering
+ 1 domain expert in sustainability science (moderator)
Group B: 4 domain experts in sustainability science
Methods
1) The subjects were assigned roles of stakeholders related to biofuel
production and policy making for it.
2) They discussed the related topics by role-playing to reach a
reasonable consensus among stakeholders.
Group A generated conceptual maps using the ontology exploration tool
and made a discussion through comparisons of the generated maps.
Group B did not use the ontology exploration tool and generated maps.
The roles of stakeholders played by subjects in the experiment
a. Industry (Sugarcane farmers, investors, Sugar processing plants, etc.)
b. Government (President's, the relevant ministry, etc.)
c. Employees (Labors union, etc.)
d. Environmental NGO
9 Oct 2012 IESD2012 19
20. Time table of the experiment
Time used Group A Group B
4 students
Group A 4 Group B
expert
in minute + 1 expert (moderator)
10 Instruction of the experiment
15 Preparation(1)[making a rough plan]
Experiment 1
20 Group discussion(1)[without the system]
Preparation(2)
15
Preparation(2) [rough planning]
35
[Each builds a map] Group discussion(2)
20 Experiment 2
[without a map]
Group discussion (2) Participate in the
20
[Discussion with maps] discussion by group B
20 Answering inquiries with wrap-up discussion
9 Oct 2012 IESD2012 20
21. Ontology explorations and
generations of maps by Group A
Methods to generate maps
To minimize the deviation of the generated maps, we restrict the
map generation command to “search path”.
The focal point (starting point): “production of biofuels”
The ending points : a couple of keywords (3 to 5) selected by the
subjects from about 120 keywords prepared in advance.
To make the maps compact and easy to interpret
The subjects delete paths which they find not interesting.
They extend paths that they want to explore further.
They got maps including only interesting
and meaningful paths according to
viewpoints of the stakeholders.
Discussion using
integrated maps displayed
on a touch-table display
9 Oct 2012 IESD2012 21
22. Result: Comparison between the
discussion done by groups A and
B
Time used Group A Group B
4 students
Group A 4 Group B
expert
in minute + 1 expert (moderator)
10 There is no significant difference of the number
Instruction of the experiment
of topics appearing the first discussion.
15 Preparation(1)[making a rough plan]
Experiment 1
20 Group discussion(1)[without the system]
Preparation(2)
15 The number of topics appearing
Preparation(2)
the second discussion [rough planning]
35
[Each builds a map] Group discussion(2)
20 Experiment 2
[without a map]
Group discussion (2) Participate in the
20
[Discussion with maps] discussion by group B
20 Answering inquiries with wrap -up discussion
9 Oct 2012 IESD2012 22
23. Discussion: Comparison between
the discussion done by groups A
and B
Usability Problem
The subjects in group A took much time to learn how to use the
system so that they did not have enough time to perform
discussion.
We had quite a few requests on improvement of the tool.
→The system needs further improvement on its usability.
Coverage of Ontology
The discussion done by group B includes concepts that are not
covered by the current ontology.
→We need extension of the ontology to cover wider and
deeper topics.
9 Oct 2012 IESD2012 23
24. Result: Discussion by Group A
through comparison of the
generated maps
Time used Group A Group B
4 students
Group A 4 Group B
expert
in minute + 1 expert (moderator)
10 Instruction of the experiment
15 Preparation(1)[making a rough plan]
Experiment 1
20 Group discussion(1)[without the system]
Preparation(2)
15
Preparation(2) [rough planning]
35
[Each builds a map] Group discussion(2)
20 Experiment 2
[without a map]
Group discussion (2) Participate in the
20
[Discussion with maps] discussion by group B
20 Answering inquiries with wrap -up discussion
9 Oct 2012 IESD2012 24
25. Result: The number of nodes included
in each map built by each subject in
group A
* The numbers of overlapping nodes indicate the how much the
stakeholders share common interests.
Number Number of overlapping nodes
of nodes in
d: Environmental
the map a: Industry b:Government c:Employees
NGO
a:Industry 110 16 21 10
b:Government 88 - 12 5
c:Employees 187 -
Employees and- 49
Environmental NGO share
d:Environmental
NGO
115 - - -
a lot of common interests.
This interpretation is supported by the result of stakeholder
analysis by an domain Sexpert [Shiroyama H, et al. 2010].
9 Oct 2012 IESD2012 25
26. Result: Distributions of
overlapping nodes
in the integrated map
In the integrated map, overlapping nodes (nodes appeared in both of
maps by different stakeholders) are show in gradation of different colors.
a: Industry ∩ c:Employees c:Employees ∩ d: Environmental NGO
Nodes in gradation of colors are Nodes in gradation of colors are
near by the center of the map. widely distributed in the map.
We can understand the differences between viewpoints of stakeholders.
26
27. Feedbacks from the subjects
The positive opinions we got from the subjects
include:
Visualization of conceptual maps is helpful to understand what
respects we are different by identifying what concepts we share
and don’t from the map.
It sometimes helps us to understand the issues better by
explicating unexpected relations or dependencies between
concepts.
It is useful for organizing my opinion to enable smooth discussion.
It is useful to clarify overlap and distinction between us objectively.
These show the feasibility and utility of the system
to some extent.
DEMO
9 Oct 2012 IESD2012 27
28. Concluding Remarks
A consensus-building supporting system based on ontology
exploration.
It generates conceptual maps through ontology exploration by the users.
Because the generated maps represent the users’ viewpoints to understand
the target domains of the ontology, it could show differences of viewpoints
through comparisons of them.
Experiment of consensus building by role-play discussion in
biofuel domain
The result shows an integrated map could well represent differences
viewpoints of several stakeholders and could help their consensus-building
through discussions using the map.
It would contribute to consensus-building on interdisciplinary domains which
consist various fields across multiple domains.
Future work
There are some rooms to improve the system because we had several
comments about its user interfaces by the subjects.
Investigations on useful viewpoints to generate conceptual maps
Application of our approach to ontology with instances and Linked Data.
9 Oct 2012 IESD2012 28
29. Acknowledgement
This research partially supported by the Environment Research and
Technology Development Fund (E-0802) of the Ministry of the
Environment, Japan and Grant-in-Aid for Scientific Research (A) 22240011.
Thank you for your attention!
HOZO with the ontology exploration tool
is available at
http://www.hozo.jp/
*The client version is available as a sub-system of Hozo.
*Web service version is also available.
9 Oct 2012 IESD2012 29