Office Automation with GIS in Urban Planning and Management...Office Automation with GIS in Urban...
Transcript of Office Automation with GIS in Urban Planning and Management...Office Automation with GIS in Urban...
Office Automation with GIS in
Urban Planning and Management
Anthony Yeh
Centre of Urban Planning and
Environmental Management
GIS Research Centre
University of Hong Kong
University of Hong Kong
GIS and Urban Planning
Urban
Planning
Spatial
Analysis &
Modelling
Non-GIS
Database/Data
data
data
Spatial Query &
Mapping
Geo-Processing
Functions
GIS
GIS
database
-GIS
commands
-macros
-GIS
commands
-macros
Data
import
export
-modelling
programs
-statistical
analysis
package
-compiled
modelling
program
GIS
database
GIS
database
-GIS commands
with modelling
functions
-macors
(a)
Loosely-Coupled Architecture
for Integrating Models with GIS
(b)
Tightly-Coupled Architecture
for Integrating Models with GIS
(c)
Fully Integrated Architecture
for Integrating Models with GIS
Integration of Models with GIS
GIS GIS GIS
Use of GIS in Urban Planning
CUPEM, The University of Hong Kong
General Administration
Development Control
Plan Making
Use of GIS in Urban Planning
General Administration
CUPEM, The University of Hong Kong
Office Automation – mapping, data retrieval
Public Enquiry – internet map, 3-D map
Use of GIS in Urban Planning
Development Control
CUPEM, The University of Hong Kong
Office Automation
Land Suitability Analysis
3-D Analysis
Use of GIS in Urban Planning
Plan Making
CUPEM, The University of Hong Kong
Office Automation - mapping
Land Suitability Analysis
Location of Facilities and Activities
Plan Generation
Plan Evaluation
Office Work
Computer Network
Human Interactions:
Not working at the same time
Not working in the same place
Multimedia information
Office Automation
System Design
Integration of Texts, Images and Maps
Textual
Databases
Graphics
Databases
ARC/INFO
001036589SQL Server
052-01
001036589
052-01
Related
Information
MapObjects
SQL Server
ARC/INFO
Dynamic Updating
Transport DB
Survey
Construction Dept
GZ
Urban Planning
Database
Construction
Database
Query
UpdatingQuery
Updating
Query
Updating
Office Automation
System
Planning
Land Use
Application
Inspection
Public Enquiry
Work Flow Management
in Development Control
Pr i v aCi t i z e Pu b l Ag e n D e p t
S u b m i
S i t
L o c a t i n
L a n d
P e r m i t t
T i t
R e g i s t e
)
L o c a
( S
R e g i s t
P r o p e
( S
L a n d
B o u n d
( S
B u i l A p p l i c a
L e g a l P e r m i
D e p t
D e p t
Priva te Cit izen or Pub lic Agen cy
Dep t. A
Subm ittin g
Sit e Lo catin g (E1 )
L and usePermit ting (E2 )
TitleRe gistering (E 3)
Loca tion (S1 )
Regist ered Prop erty (S3 )
L and useBoun dary
Build ing App licatio n(s)
Le gal Permit or No tice
Age nts Even ts Sta tes
Ur ba n Pla nning a nd La nd Management Organiza ti onAppli cant
Be haviora l Aspects Structur al Aspec ts
De pt. C
Dep t. B
Work Flow Management
Public Enquiry - Internet
Office Automation with
Case-Based Reasoning and GIS
CUPEM, The University of Hong Kong
Introduction (1)
Assumptions of KBS (rule-based or model-based)
(Han and Kim, 1990):
• the problem is clearly specifiable and well-
bounded;
• the ‘deep knowledge’, i.e. the relations between
the factors or elements of the problem, are
known and can be expressed explicitly;
• problem solving methods can be articulated; and
• experts agree on solutions, i.e. the cause-effect
relationships are definitely defined.
Introduction (2)
• Unfortunately, these requirements are seldom
met in planning.
“It is very difficult to find an urban planning problem
which is well-bound, well understood, and takes a few
minutes to a few hours to solve” (Han and Kim, 1990).
CBR and Its Advantages to
Urban Planning
Case-Based Reasoning (CBR) is a method or
technique of KBS which uses the previous
similar case(s) to help to solve a current new
problem (Kolodner, 1993, pp.4).
CBR can be typically represented by a schematic
cycle comprising four REs :
RETRIEVE the most similar case(s);
REUSE the case(s) to attempt to solve the problem;
REVISE the proposed solution if necessary, and
RETAIN the new solution as a part of a new case.
CUPEM, The University of Hong Kong
The CBR Cycle
RETRIEVE
RETAIN
REVISE
REUSE
Problem
Case-Base
Library
Confirmed
Solution
Proposed
Solution
Advantages of Case-Based
Systems CBS (1) store knowledge in the form of concrete stories
(cases) rather than abstract rules or models.
CBR provides a good method to involve
‘informal’ information’, e.g. personal judgements,
hunches, intuition, hearsay and personal
experiences (Han and Kim, 1990).
CBR can provide raw but comprehensive and
original information to the planner, and leave
space for the intelligence of planners to
recognize and understand the problem.
Advantages of Case-Based
Systems CBS (2)
solution to a problem can be conveniently saved
in the case-base and the system’s knowledge
increases.
Learning method is very simple and can be
performed by experts themselves during working
process without knowledge engineers’ help.
CBR will not just tell the planner what he or she
should do but present previously handled similar
problems to the planner
Applying CBR in Processing
Planning Applications• Processing planning applications is a daily work
in the Planning Department of the Hong Kong
Government.
• Make recommendations to the Town Planning
Board
The Integration of CBS
(ESTEEM) and GIS (ArcView)• CBS software - ESTEEM
• GIS software - ArcView
integrated through DDE of Windows
Representation of a
Planning Application Case
A planning application case contains three parts :
graphics (map, picture, charts) (handled by
ArcView)
feature values (dbf file, handled by ESTEEM)
comments (txt file)
– valuable tips about this case
– list of cases which are relevant to this case
CUPEM, The University of Hong Kong
Integration of CBS and GIS
User Interface
GIS(ArcView)
CBS(ESTEEM)
Graphics CommentsDescribing
Features
User Inputs System Outputs
CBR Result(Information about the Retrieved Cases)
Descriptions of
the New Application
Identifying
Codes
Contents of
the Specified CasesMatching
DirectoriesInformation about
the Retrieved Cases
Shape Files Text FilesFeature
Values
Case-Base Library
Main Retrieval Functions of
the System (Office Automation)
• Spatial Retrieval
• Nonspatial Retrieval
CUPEM, The University of Hong Kong
An Example of a Planning Application Case
An Example of Index Map for Spatial Retrieval
Nearest Neighbour Algorithm
• assess similarity between stored cases and the new
problem based on matching a weighted sum of
features (Kolodner, 1993, pp. 354; Watson, 1995).
Case Retrieval : Similarity Index
Nearest Neighbour Algorithm
assess similarity between stored cases and the new
problem based on matching a weighted sum of features
(Kolodner, 1993, pp. 354; Watson, 1995). The algorithm
is represented as
wi = importance of dimension i
sim = similarity function for comparing
feature fiR and fi
I
fiR and fi
I = values for feature fi in the input (I) and
retrieved cases (R), respectively
(Kolodner, 1993, pp. 355).
w sim f f
w
i iI
iR
i
n
ii
n
( , )1
1
Non-Spatial Retrieval
Similarity-Based Retrieval
If more than one similar case are found, the result
will be listed in a table of ArcView:
• Identification Code (Appnum)
• Approval Probability (Approval)
• Similarity Score (Similarity)
Approval Probability = 0 to 1
0 = not approved
1 = approved
List of Similar Cases
• It will calculate the approval probability of the new
application based on previous similar cases.
• The New Approval (NA) probability is calculated with the
following formula:
Evaluation of a New Case
NA = proposed probability of approving the new application
ai = approval probability of each retrieved case
si = similarity of each retrieved case
NAa s s 2
s2
i i i
i
- +
((( ) ) )0 5 0 5. .
Evaluation of a Planning Application
Case Library
Adjustment by
the Experts
Proposed Approval Score
of the new application
Result of the
application
Evaluation Documents
Application
Documents
Deliberation of
Town Planning Board
Proposed Approval Score
of the new application
Operations of
CBR Engine
Information for matching
Features of the
new application
Modified Approval Scoreof the new application
... ...
Data of the new case
Similar Case1
(Approval Score1)
Similar Case2
(Approval Score2)
Similar Casen
(Approval Scoren)New Case
RETRIEVE
REVISE
REUSE
RETAIN
Case-Based Reasoning and GIS
An Office Automation System
(text-graphics integration)
An Expert System for Experts
Corporate Memory System
Ensure Consistent Decision
CUPEM, The University of Hong Kong
Discussion
• The integration of CBR with other reasoning
engine to deal with situation without cases.
CBR can be integrated with other reasoning
engines, e.g. rule-based reasoning (RBR) engine
and model-based reasoning (MBR) engine, to build
a more powerful KBS (Bartsch-Sporl, 1995;
Burstein, etc., 1996; Kolodner, 1993, pp.92-96).
• Maintenance of cases - time decay function