Spatial databases: Introduction

1
Spatial databasesIntroduction

• Geog 495 GIS database design

2
Outlines

• Decoding the acronym GIS
• GIScience view on spatial database
• DBMS view on spatial database
• Review questions

3
1. Decoding the acronym GIS

• GI Systems
• GI Sciences
• GI Services
• Evolution of GIS

4
GI Systems

• Geographic information system (GIS) is a system
  for input, storage, manipulation, and output of
  geographic information (NCGIA CC)
• Database system designed to handle geographically
  referenced data
• Composed of data, software, hardware, people and
  procedure (ESRI)

5
GI Sciences

• The science behind the technology
• Aimed at enhancing knowledge of geographic
  concepts and their computational implementations
• seeks to redefine geographic concepts and their
  use in the context of geographic information
  systems
• Dedicated to the development and use of theories,
  methods, technology, and data for understanding
  geographic processes, relationships, and pattern
  (UCGIS 1994)

6
GI Services

• Gradual shift from centralized GIS to distributed
  GIS
• Examples include location-based service,
  Web-mapping, Web-based planning support system
• GIService is miniature, mobile, public, and
  task-specific
• If GISystem is data-centered, GIService is
  person-centered

7
Evolution of GIS

• GIScience view
• Database view

GIServices
GISystem
technology
GIScience, Database
research
8
2. GIScience ViewOvercoming limitations of
existing GISystem

• Challenges
• Needs
• Subjects of GIScience

9
Challenges

• Because progress has historically relied on a
  fragmented gathering of approaches inherited from
  cartography, imposed by hardware, or borrowed
  from other computer-related fields, we are faced
  with the current situation in which increased
  functionality has characteristically been
  accompanied by increased conceptual complexity,
  making GIS progressively more nonintuitive for
  the user.

Representations of space and time by Donna J.
Peuquet, 2002
10
Needs

• Need better ways to represent, understand,
  manage, and communicate our natural world

11
Subjects of GIScience

• How people think about geographical space and
  time
• Ontology of geographic kind
• How to translate human conceptualizations into
  formalisms that allow these processes to be
  repetitively consistent
• Formalism of spatial language
• How to make people interact more naturally with
  information systems
• System design

Egenhofer et al, 1999
12

• Further readings on GIScience
• Goodchild, M.F., 1992, Geographical information
  science. International Journal of Geographical
  Information Systems 6(1) 31-45
• Goodchild M.F., 1997,What is Geographic
  Information Science?, NCGIA Core Curriculum Unit
  2
• http//www.ncgia.ucsb.edu/giscc/units/u002/
• Mark D., 1999, Geographic Information Science
  Critical issues in an emerging cross-disciplinary
  research domain, workshop on Geographic
  Information Science and Geospatial Activities at
  NSF
• http//www.geog.buffalo.edu/ncgia/GIScienceReport
  .html

13
3. Database System ViewOvercoming limitations of
existing DBMS

• Challenges
• Needs
• Evolutions of DB systems
• GIS architecture
• SDBMS architecture
• Subjects of spatial databases

14
Challenges

• Previous DBMS is not well accommodated into
  geographic concepts as most of commercial DBMS
  are designed to handle attribute data
• Whats special about geographic? ? go to the next
  slide

15
Properties of geographic data

• Has location
• Location is a special kind of key (i.e. list of
  values) how is it handled?
• Multidimensional
• Directional, topological relationships, how is it
  formalized
• Scale-dependent
• Spatial versus Geographic
• Its occurrence is spatially autocorrelated
• Toblers first law of geography
• Not well captured by precise description
• Uncertainty should be formalized
• Some geographic phenomena are continuous
• Object-view wouldnt fit well
• Some geographic phenomenon is closely associated
  with temporal changes
• event, process, moving object

Geographic data need special treatment indeed!
16
Needs

• We need more constructs to handle spatial
  information in order to reduce the semantic gap
  between the users view of spatial data and the
  database implementation

Spatial Database a Tour by Shashi
Shekhar and Sanjay Chawla, 2003
17
Evolution of DB system

• File systems
• Network DBMS
• Hierarchical DBMS
• Relational DBMS
• Object-oriented DBMS
• Object-relational DBMS

18
What is object-Relational DBMS?

• Add OO-ness to tables
• All persistent (database) information is still in
  tables, but some of the tabular entries can have
  richer data structure, that is ADTs
• ORDBMS supports an extended form of SQL
• Potential for mapping spatial concepts
• For example, Oracle 8i implements spatial data
  types and spatial operators

19
GIS architecture

• Built upon File system Relational DB
• Spatial data is stored in file system
• Attribute data is stored in tables
• Separation between non-spatial and spatial data
• Specific module for spatial data management
• Also called georelational model
• Examples Arc/Info, MGE, TiGRis (Intergraph)

20
GIS architecture

• Built upon relational DB
• Spatial/non-spatial data are stored in tables
• Storing spatial data in table is tedious
• Can hinge on SQL
• Violate data independence principle
• Bad performance
• Difficulty in defining new (spatial) types

21
GIS architecture

• Built upon object-oriented DB
• Can define user-defined data type (e.g. ArcGIS
  geodatabase has different data models such as
  hydrology, network, land parcel, and so on)
• Can define user-defined operations
• Can inherit properties and operations from
  superclass

22
How to map land parcel data?

• In relational database
• e.g. arc/info coverage, arcview shapefile
• spatial data is stored in a file and its linked
  to attribute data through common attributes
• In object-oriented database
• e.g. arcgis geodatabases
• you can inherit properties from object,
  featureclass, polyline and call the methods
  area() defined in polyline

23
Spatial DBMS

• Special kind of DBMS that are specifically
  designed to handle spatial data
• Usually seen as middleware (e.g. ArcSDE)
• Can be implemented in either thick or thin client
  (e.g. CGI versus Java)
• Have different capabilities depending on which
  database models
• Usually support spatial indexing, efficient
  algorithms for spatial operations, and
  domain-specific rules for query optimization

24
Architecture of SDBMS
25
Subjects of spatial databases

• Spatial taxonomy
• Data model
• Query language
• Query processing
• File organization and indices
• Query optimization
• Data mining

26
Review questions

• Discuss the differences between spatial and
  nonspatial data
• How can object-relational databases be used to
  implement an SDBMS?
• Compare and contrast
• GIS vs. SDBMS
• OODBMS vs. ORDBMS
• GI Systems vs. GI Services
• Querying vs. Data mining