PROBLEM SOLVING METHODOLOGY AND PROJECT DESIGN
Note: Content adapted from chapter 12 of: Heywood, I., Cornelius,
S. and S. Carver, 1998. An Introduction to Geographical Information
Systems, Addison Wesley Longman Ltd., New York.
PROBLEM-SOLVING APPROACH
Soft Systems
Unstructured problems
Multiple users with
different needs all using the same system, e.g. a GIS for people
choosing a home
Hard Systems
Structured problems
Model reality or well
defined scenarios, e.g. a GIS showing where homes for sale are located
PHASES OF HARD SYSTEMS DESIGN
Lexical Phase
1. Define the problem
Background of problem
Review of similar/associated
work
Needs Analysis
Problem statement which clearly
identifies the question to be answered.
2. Define the boundaries (scope) of the problem
Spatial extent
Temporal extent
Major Assumptions already established
3. Define Entities
Real world features of interest
including spatial/temporal characteristics
Data sources + georeferencing
(include error levels)
Entity spatial definition (point,
line, polygon, volume)
4. Define Entity State
Specify attributes of entities
(define tables)
Data sources + available error
levels
Attribute measurement
level and units
Parsing Phase
1. Conceptual Data Model
Identify if entities are raster
or vector data structures and group as data layers or objects.
Identify topological features
of data within data layers or between data objects.
Grid resolution or fuzzy tolerance
Specify linkages between attribute
tables
2. Defining Relationships between Groups of Entities
Permissible entity relationships
Linkages between attributes
and entities
Linkages between attribute tables
3. Physical Data Model
Identify digital data format
of input spatial data and attributes
Specify application specific
data models within GIS
Specify georeferencing
standards within GIS
Modeling Phase
1. Data Flow
Use map algebra and non-system
specific terms to specify data flow.
Clearly indicate interfaces
to analytical models external to GIS
2. Analytical Models
System specific flowchart.
Define methods for analytical
modeling (e.g. regression, clustering etc.)
Perform analysis external to
GIS which only supplies inputs.
Analysis Phase
Present modeling results: perform
GIS analysis
Validate: use real world
validation or perform error analysis
Document: identify what
was accomplished in terms of initial problem and what can be improved
upon.
DATA DICTIONARY
All projects are
required to create a Data Dictionary for each data layer they create (either
through analysis on existing GIS data or imported data). For the purposes of
this course you can use the data dictionary format established for the Humber
River watershed database (see
example), or the ArcGIS Metadata format found in ArcCatalog. All groups will
be required to produce these data dictionaries as part of their final report.
If you are making up a new project you may want to look at the National
Topographic Database Data Dictionaries.
ASSESSMENT WILL BE BASED ON:
completeness
presentation (10 minute web
presentation + 5 minute questions)
conciseness (proposal sticks
to lexical phase)
extent to which proposal is
realistic
amount of research put into
lexical phase
|