Geog 495: GIS Database Design

9/29/05


Data vs. Information

 

Example A

A pile of sales inventories for item a and b à you have found out they are increasing, and the gap between two items is widening by graphing trends

 

 

Example B

A collection of satellite images à you compare the urban land use from year 1973 to year 2000, and found out that it has expanded (i.e., urban sprawl?)

 

http://www.grid.unep.ch/activities/global_change/atlas/images/LasVegas.jpg

 

Processing data helps you to reveal the meaning, or even help you identify phenomenon you couldn’t see otherwise. What’s the role of database design on this transformation from data into information?


Data vs. database

 

The difference

can be viewed from set theory: which is bigger? Actually more than that

can be viewed from end user’s view: does it serve a specific purpose of end users?

Both requirements should be met to determine which is database

 


Database vs. database management system (DBMS)

 

 

DBMS is a collection of programs that manages the database structure and controls access to the data stored in the database; allows for sharing data among multiple applications or users (e.g., Oracle, Informix, SQL Server)

 

Spatial DBMS: specialized DBMS to meet the requirements of spatial data (SC1:10)

e.g., ArcSDE, Spatial Data Blade, Spatial Data Cartridge

 


DBMS vs. database system

 

Database system is composed of five main components

(1) hardware (2) software (3) people (4) procedures (5) data

 

 

 


Need for database system

 

Limitations of file systems: data redundancy, data dependency

 

 

Data redundancy: different and conflicting versions of the same data appear in different places; lead to data inconsistency

 

Data independency: separation of data from programs that use the data

 


Different types of database

·        Number of users

Ø      Single-user

Ø      Multiuser

Ø      Workgroup

Ø      Enterprise

·        Database site location

Ø      Centralized

Ø      Distributed

·        Database use

Ø      Transactional (production) database

Ø      Data warehouse database


Why database design is important

 

DB design is like blue print for building house

Good applications can't overcome bad database designs


Lab

Open files P:\geog495aut05\Ch01_Problems.mdb

Open PROB_1_01 in the table

 

1. How many records does the file contain, and how many fields are there per record?

 

2. What problem would you encounter if you wanted to produce a listing by city? How would you solve this problem by altering the file structure?

 

Open PROB_1_05 in the table

 

3. Identify and discuss the serious data redundancy problems exhibited by the file structure.

 

4. Identify the different data sources in the file you examined.