SYSTEMS INTEGRATION SYS5160 Instructor: Ali Abbas Term Project: Flexible Support for Spatial Decision-Making By: Mohamed Ebada & Peter Timusk Date: February 8 th 2008
May 06, 2015
SYSTEMS INTEGRATION SYS5160
Instructor: Ali Abbas
Term Project:
Flexible Support for Spatial Decision-Making
By: Mohamed Ebada & Peter Timusk
Date: February 8th 2008
Flexible Spatial Decision Making Support Systems:
the background to the paper
The paper selected as a base for the project is:
Flexible Support for Spatial Decision Making
Shan Gao, John Paynter, and David Sundaram Proceedings of the 37th Hawaii International Conference on System Sciences – 2004
We take the approach of a literature review by presenting tonight a brief summary of the references that the authors used in their paper.
We tie in structures we have looked at in the engineering life cycle but focus on the structures of the processes that the authors use for their decision making support system. In particular, we trace the roots of the structures they used. We try to answer why did the authors use the structure(s) they did and where did they get their ideas, concepts and motivations.
We also cover a longer version of the author’s paper we found that was a whole book chapter at the end of this presentation. That presentation will present more examples of their decision making system at work. Here will offer our critique and suggestions for how to now improve their system even further.
We can see that the authors developed existing theory and moved from this to where they arrived.
Now the references ---->
The author’s references
• [1] Densham, P. J. (1991). Spatial Decision Support Systems in: Maguire, D.J., Goodchild, M.F. und Rhind, D.W. (Ed.): Geographical Information Systems: Principles and Applications. Longman, Burnt Mill (UK), pp. 403-412.
• [2] Geoffrion, A.M. (1987). An Introduction to Structured Modeling. Management Science 33(5): 547-588.
• [3] Malczewski, J. (1998). Spatial Multi-Criteria Decision Analysis in Thill, J-C (Ed.): Spatial Multi-Criteria Decision-making and Analysis: A Geographic Information Sciences Approach. Brookfield, Ashgate: pp. 11-48.
• [4] Moloney, T., Lea A.C. and Kowalchek, C. (1993). Manufacturing and Packaged Goods in Profiting from a Geographical Information System, GIS World Books Inc, Fort Collins.
• [5] Peterson, K. (1998). Development of Spatial Decision Support Systems for Residential Real Estate. Journal of Housing Research 9(1): Fannie Mae Foundation.
• [6] Post, Elizabeth (2000). Jade for Developers, 2nd Edition. Auckland, NZ, Pearson Education.
• [7] Simon, H.A. (1960). The New Science of Management Decision. New York, Harper and Row.
[1] Densham, P. J. (1991). Spatial Decision Support Systems in: Maguire, D.J., Goodchild, M.F. und Rhind, D.W. (Ed.): Geographical Information Systems: Principles and Applications. Longman, Burnt Mill (UK), pp. 403-412.
Spatial Decision Support System[1]
• Spatial problems are ill or semi structured• An example is: locating a bank branches.• Geoprocessing systems must support a
decision research process that explore the problem to increase the level of understanding and investigate the possible solutions by generating and evaluating alternatives.
• Decision making proceeds by multiple paths involving different routes
Additional characteristics for SDSS:
• Mechanisms for the input of spatial data
• Representation of the complex spatial relations and structures
• Analytical techniques that are unique to both spatial and geographical analysis (statistics)
• Provide output in a variety of spatial forms including maps and other types
Architecture of SDSS:
• A set of five integrated software modules– Database management system– Model base management system– Display generator– Report generator– User interface
Geoffrion’s DSS characteristics (1983):
• Designed to solve ill-structured problems where the problem can not be precisely defined
• Powerful and easy to use user interface• Enable users to combine analytical models and data
in a flexible manner• Can generate a series of feasible alternatives to help
the user explore the solution space• Easily adapted to provide new capabilities as the
needs of the user evolves• Allow problem solving to be both interactive and
recursive
[2] Geoffrion, A.M. (1987). An Introduction to Structured Modelling. Management Science 33(5): 547-588.
Total life-cycle associated with model-based analysis and systems
•Determine requirements
•Design
•Build
•Test
•Use
•Revise
•Maintain
•Document
•Explain
• analyze results
•report findings
•Evolve
[2] Geoffrion, A.M. (1987). An Introduction to Structured Modelling. Management Science 33(5): 547-588. Continued.
The authors evaluate the problems with model based software based to a large extent on the criteria that Geoffrion sets for a modeling software for structured modeling. Here are his eight criteria written in 1987.a) A rigorous and coherent frame work for modeling…b)Independence of model representation and model solution,…c)Sufficient generality…d)Usefulness…e)representational independencef) Desktop implementationg) Integrated facilitiesh) Immediate expression evaluation in the tradition of desktop spreadsheet software.p. 549
a) Rigorous
A rigorous and coherent frame work for modeling based on a single model representation format suitable for managerial communication, mathematical use, and direct computer execution
b) Independence
independence of model representation and model solution, with model interface standards to facilitate building a library of models and easily accessed solvers for retrieval, systems of simultaneous equations, optimization, and other important manipulations.
c) generality
Sufficient generality to encompass most of the modeling paradigms that MS/OR and kindred model-based fields have developed for organizing the complexity of reality ( activity analysis, decision trees, flow networks, graphs, markov chains, queuing systems.
d) Usefulness
Usefulness for most phases of the entire life-cycle associated with model-based work.
e) representational independence
representational independence of general model structure and detailed data needed to describe specific model instances
f) Desktop implementation
Desktop implementation with a modern users interface (e.g., visually interactive, directly manipulative, syntactically humane, and with liberal use of graphics and tables)
g) Integrated facilities
Integrated facilities for data management and ad hoc query in the tradition of database systems
h) Immediate expression
Immediate expression evaluation in the tradition of desktop spreadsheet software.
[3] Malczewski, J. (1998). Spatial Multi-Criteria Decision Analysis in Thill, J-C (Ed.): Spatial Multi-Criteria Decision-making and Analysis: A Geographic Information Sciences Approach. Brookfield, Ashgate: pp. 11-48.
This reference is to the beginning of a book where the book covers data representation of a GIS. They do not reference the entire book. Later chapters cover spatial multi-criteria decision making but for this presentation we will not look deeply at this book.
The later chapters have a taxonomy of decision support systems such as the taxonomy used by our authors.
Suffice it to say that this book is also about the same problem area but it seems the authors have only used the beginning of this reference where basic GIS and problems solving with GIS are introduced.
[4] Moloney, T., Lea A.C. and Kowalchek, C. (1993). Manufacturing and Packaged Goods in Profiting from a Geographical Information System, GIS World Books Inc, Fort Collins.
Profiting from a GIS [4]
• GIS allows business professionals to better understand the effect of geography on business performance and profitability
• Greater than 90% of data are geographic in nature• For marketing to be effective and efficient, geographical
data need to be analyzed• GIS helps Manufacturers and packaged goods
companies in their decision-making process in marketing, advertising and sales management
• Case studies:– Effective deployment of sales effort– Brand targeting, private label– Rationalization of automobile dealerships
[5] Peterson, K. (1998). Development of Spatial Decision Support Systems for Residential Real Estate. Journal of Housing Research 9(1): Fannie Mae
Foundation.
The IT background paper with justifications for object oriented programming and also matching Geoffrion for arguments for flexible systems and decision support systems that have multiple applications.
Specifically in a real estate journal
We will look at this paper a little more but not cover the technical computer details.
[5] Peterson, K. (1998). Development of Spatial Decision Support Systems for Residential Real Estate. Journal of Housing Research 9(1): Fannie Mae Foundation.
Functional areas for spatial decisions making in real estate
1. Planning, strategy formulation, and research
2. Construction and development
3. Finance, including construction lending, equity investment, mortgage finance, public and private venturing, and property taxation and assessment
4. Property management
5. Risk management
6. Marketing
7. Regulatory compliance
p. 136
[5] Peterson, K. (1998). Development of Spatial Decision Support Systems for Residential Real Estate. Journal of Housing Research 9(1): Fannie Mae
Foundation.
Benefits for creating and implementing an IT architecture (Peterson quoting Rosser, 1995)
• The ability to achive interoperability--that is, different systems working together, especially in sharing data.
• The ability to speed the implementation of new systems by having many choices already settled and the skills and learning in place
• Lower costs due to the reduction of support efforts as the number of products and processes is reduced
• The general upgrading of system quality and increased ability to make modifications more easily in the future.
• Communication of a common direction throughout IT and end-user departments, a by-product of going through the architecture planning process
• In the effort to reduce the total number of components and processes employed, an architecture may forestall the crsis of complexity that is projected for networked computing as more and more demands are made and expectations rise.
p. 141
[7] Simon, H.A. (1960). The New Science of Management Decision. New York, Harper and Row.
Pioneer in automation and decision making. Early operations research.We did find this book yet. We may include it in our written report. Other books he wrote suggested the new computer age was going to help support decision making.
French available translation in Library.
Gao, S and Sundaram, D. Flexible Spatial Decision-Making and Support : Processes and Systems in Hilton, B. N. ed. Emerging Spatial Information Systems and Applications,
Idea Group, Hershey, PA .: pp. 153-183.
A longer version of the paper as a chapter in a book.
FSDSS• FSDSS is an interactive computer-based
systems that is flexibly and dynamically integrate spatial and non-spatial data, models, and solvers to explore, transform and evaluate multicriteria spatial decision scenarios for solving complex spatial problems.
Sample sessions with the FSDSS
• The FSDSS system proposed by Shan Gao, John Paynter, and David Sundaram were used to solve three spatial problems:– Design of a running track– Purchase of a house– Trace the spread of a disease
Design of a running track
• Plot a number of paths by clicking points on the map
• Paths are saved in the database as running path 1 and running path 2
• Consider:– Drinking points– Toilets– distance
Design of a running track
The two alternative paths for running event
Spatial layer manager
Spatial layers (toilet and drink point layer)
Implementation of spatial modeling
• Drinking points and Toilets are stored in the database as object.
• A collection of all Drinking points and all the Toilets would form a Drinking points layer Toilets layer
• Loading the two layer would show the objects on the layer on the map and relate them to the paths to create a theme that can be saved or changed
• the distance of every path can be calculated using the distance solver
• These scenarios are compared
Purchase of a house• The analysis contain spatial and non spatial parts• Spatial analysis include environment factors like
school zone and distance to main facilities factors like distance to school, hospital and public transportation.
• Non spatial analysis include price factor and functions factors like number of rooms and a availability of a garage
• Value system:– Quality criteria: construction material and year built– Economic criteria: market price– Location criteria: property accessibility
• User select a target area• Load relevant decision making components• Spatial and non spatial scenario are
developed separately and then integrated into combined one scenario
• Use goal seeking and sensitivity analysis to help chose between alternatives
Purchase of a house
Simple spatial scenario creation
Simple non-spatial scenario creation
Integration of non-spatial with spatial scenarios
Integration of spatial with non-spatial scenarios
Multi-attribute spatial scenario creation
Multi-criteria spatial scenarios
Trace the spread of a disease
• Identify the spread of the SARS epidemic in Hong Kong
• Understand, analyze and manage geographically distributed data using spatial functions to discover spatial patterns
• Obtain a geographical distribution of SARS cases and the population at any time
Trace the spread of a disease
Case occurrences and population density
Time series presentation of SARS spread period I
Time series presentation of SARS spread 2
Infected buildings, school, and others
The steps in the authors’ process
• Problem Identification
• Problem Modeling
• Scenario Development
• Scenario Integration and Instantiation
• Scenario Execution
• Scenario Evaluation
• Decision-Making
The steps in the authors’ processcompared to steps in Hall’s
decision making process
• Problem Identification
• Problem Modeling
• Scenario Development
• Scenario Integration and Instantiation
• Scenario Execution
• Scenario Evaluation
• Decision-Making
•Problem definition
•Value System
•Options
•Options Analysis
•Optimization
•Decision-Making
•Preparing for next phase
Conclusion
The authors used a number of sources to identify a problem, become familiar with the problems domain, find recommended solutions within that domain, look at a long term view of the problem with many years of research span.
Therefore they “stood on the shoulders of giants” to paraphrase Einstein.
References: [1] Densham, P. J. (1991). “Spatial Decision Support Systems” in: Maguire, D.J., Goodchild,
M.F. und Rhind, D.W. (Ed.): Geographical Information Systems: Principles and Applications. Longman, Burnt Mill (UK), pp. 403-412.
[2] Geoffrion, A.M. (1987). “An Introduction to Structured Modelling”. Management Science
33(5): 547-588. [3] Malczewski, J. (1998). “Spatial Multi-Criteria Decision Analysis” in Thill, J-C (Ed.): Spatial
Multi-Criteria Decision-making and Analysis: A Geographic Information Sciences Approach. Brookfield, Ashgate: pp. 11 48.
[4] Moloney, T., Lea A.C. and Kowalchek, C. (1993). Manufacturing and Packaged Goods in
Profiting from a Geographical Information System, GIS World Books Inc, Fort Collins. [5] Peterson, K. (1998). “Development of Spatial Decision Support Systems for Residential Real
Estate”. Journal of Housing Research 9(1): Fannie Mae Foundation. [6] Gao, S and Sundaram, D. Flexible Spatial Decision-Making and Support : Processes
and Systems in Hilton, B. N. ed. Emerging Spatial Information Systems and Applications, Idea Group, Hershey, PA .: pp. 153-183.
[7] Simon, H.A. (1960). The New Science of Management Decision. New York, Harper and Row.
Graphics on Slide 2 from wikipedia <http://en.wikipedia.org/wiki/Geographic_information_systems> (accessed February, 2008)
The endfini
Questions?
56