From Event to Action: Accelerate Your Decision Making with Real-Time Automation
How to integrate the results of the Country Reports in the Semantic Wiki
1. How to integrate the results of
the Country Reports in the Semantic Wiki
Silvana R. Nobre
Atrium Forest Consulting, Brasil
silvana@atriumforest.com
Luiz Carlos E. Rodriguez
University of São Paulo, Brasil
lcer@usp.br
Atrium – Forest Consulting
Zvolen 2012
2. How to integrate the results of the
Country Reports in the Semantic Wiki
1. Actual situation & objectives
2. Knowledge representation basic concepts
3. Models & Examples
4. Attribute Values
5. What is ahead?
Atrium – Forest Consulting
Zvolen 2012
3. 1. Actual situation & Objectives
a. We have collected data about the use of DSS on Forest
Planning in each Country.
b. We have interpreted data and their relations
c.
Those data make sense in compiled reports
d. We have similar reports for each Country
e. We have read and analyzed CR under some perspective
Context
Wisdom
Principles, explanations
Knowledge
patterns
Information
We are here!!
relations
Data
Understanding
Atrium – Forest Consulting
Zvolen 2012
4. 1. Actual situation & Objectives
a. Guidelines to development
b. Guidelines to use of DSS
c.
Lessons learned…
We want to get there!!!
d. …
Context
Wisdom
Principles, explanations
Knowledge
patterns
Information
relations
Data
Understanding
Atrium – Forest Consulting
Zvolen 2012
5. 1. Actual situation & Objectives
1. Understand patterns on Country Reports
•
We have already done a great effort to it!!
•
How far have we got on standardization?
2. Do knowledge model
3. Agree on a single “place” to store our knowledge base
4. Bring to the knowledge base what is already done.
5. Create a Knowledge base
Knowledge
People
6. Semantic wiki will make the knowledge base available
7. Look for explanations, principles, guidelines…
Value!!!
Technology
Atrium – Forest Consulting
Zvolen 2012
6. 2. Knowledge representation basic concepts
Knowledge from experience
1. Objects & Class of Objects that participated the experiences we want to
transform in knowledge.
2. We must choose which objects from the real world we are going to represent
3. We must describe those objects
4. Describe an object means “say what are the attributes of each object”
5. Each attribute has his own set of possible values
1. What are our relevant objects?
Answers are inside the CR
2. What are their common attributes?
3. What are the attributes possible value?
Atrium – Forest Consulting
Zvolen 2012
7. 2. Knowledge representation basic concepts
Knowledge from experience
There will be an agent (a robot, or a semantic wiki ) ready and prepared to
read information and present as knowledge to humans.
The agent deals with (object, attribute, value) vectors.
Semantic Wiki
Structured
&
Interpreted
Information
Atrium – Forest Consulting
Zvolen 2012
8. 3. Model & Examples
Objects of the real world
Level of object description can be:
•
•
•
CR + Problem Type
DSS + Problem Type
CS + Problem Type
Important task: common
classification for Problem Type
•
•
Problem Type #1
Atrium – Forest Consulting
•
•
•
•
Temporal scale: strategic;
Spatial context: spatial with no
neighborhood interrelation;
Spatial Scale: stand level;
Single Decision making
Single objective;
Function: wood products
Zvolen 2012
9. 3. Model & Examples
Describing Objects
Examples on “Object, Attribute, Value” representation.
•
(Austria Problem Type 1; Model; “Forest Species Conversion”)
•
(Austria Problem Type 1; KM Technique; “Relational Database”)
•
(Austria Problem Type 1; Volume estimation; “process -based”)
•
(Austria Problem Type 1; Modeling Techniques; “Multi-criteria Analysis”)
heritage:
•
•
A DSS+PT can inherit an attribute from a CR+PT
CS + PT can inherit an attribute from a CR+PT
Example on heritage:
•
ClimChalp DSS inherits the attribute “Forest Species conversion” from
Austria Problem Type 1
Atrium – Forest Consulting
Zvolen 2012
10. 3. Model & Examples
General Model
Atrium – Forest Consulting
Zvolen 2012
11. 3. Model & Examples
Model Application: Graph Examples
Atrium – Forest Consulting
Zvolen 2012
12. 4. Attribute Values
1. It is a potential problem
2. Some of them we already know the set of possible values
3. We must avoid replication: same meaning, different words
4. Suggestion: concentrate the job in a team (STSM)
5. Attach two attribute values of a same attribute class to an single object
6. Classical examples:
Models
Modeling Techniques
•
•
•
•
•
•
Linear programming
Goal Programing
Heuristics
Multi-criteria analysis
Dynamic Programing
…..
Atrium – Forest Consulting
•
•
•
•
•
•
•
Harvest Scheduling
Species Conversion
Transportation
Daily truck dispatch
Road design optimization
Genetic Material Allocation
….
Zvolen 2012
13. 6. What is ahead?
1. Understand patterns on Country Reports
•
We have already done a great effort to it!!
•
How far have we got on standardization?
2. Validate a knowledge model
3. Agree on a single “place” to store our knowledge base
4. Bring to the knowledge base what is already done.
5. Create a unique Knowledge base
6. Semantic wiki will make the knowledge base available
7. Look for explanations, principles, lessons learned, guidelines…
Atrium – Forest Consulting
Zvolen 2012