Transcript Database Introduction - YSU Computer Science & Information Systems

Database Management
Exploring the Territory
Database vs Flat Files
• Flat Files
– Characters-fields-records-files
• Files are not designed to work together
– Each file is an independent entity
– No design rules followed
– Updates can be hairy
• Databases
– Files designed to work together
• Joins supported – fields from two or more files
• Only redundancy is in linking (foreign) fields
• Updates should result in few if any errors
Database Models
• Hierarchical
– Resembles DOS tree structure
• Network
– Hierarchical plus relationships that move two directions
• Relational
– Offers set of underlying principles
– Too much hype, too many fads, everything is relational,
lack of standards
• Object Oriented
– Objects, classes, etc.
Our Goal
• Provide a correct up-to-date understanding
of and appreciation for the most practical
aspects of crucial database issues.
• Clarify concepts, principles, and techniques
that trouble vendors and users
• Specific suggestions on how to deal with
SQL implementations
Fads
• Vendors exploit ignorance
– Obscure serious product deficiencies
• Panaceas for Database Management (that
are/were not)
–
–
–
–
Internet
Virtual Machines
PC
SQL
ASP
ColdFusion
Complex Data Types
Real Database Problem
• Failure of vendors and users to educate
themselves
• Failure to rely on scientific foundation
• Use of ad hoc cookbook approach
• Self taught individuals who become DBAs
Fundamental Issues
• Database work needs to be approached
scientifically
• No amount of expertise in a product or
platform is sufficient in and of itself to
address database design issues
• Consequences of wrong approach are severe
• Designer needs knowledge and
understanding of fundamentals
Solution
• Database education – not product education
• Identify the true database issues
–
–
–
–
–
–
–
Unstructured data and complex date types
Business rules and integrity enforcement
Keys
Quota queries
Duplicates
Redundancy
Normalization & denormalization
Entity subtypes and supertypes
Data Hierarchies and recursive queries
Your Text
• Demonstrates how impractical and costly
ignoring fundamentals can be
• Examples are from actual database projects
and include specific SQL or other solutions
• Material is reasonably understandable to
non-technical reader
• Compact – each chapter is 15 pages or less
Outcomes
• Understand central issues in database management
• Avoid costly misconceptions and fallacies
• Appreciate correct general solutions to core
problems
• Assess if DBMS software offers a solution
• Overcome, work around, or minimize
consequences when products do not offer
solutions
Data Modeling Questions
•
•
•
•
What information is in the system?
How is the information organized?
How can users obtain information they need
Questions to be answered
– What entities do we model
– What relationships exist between the entities
– What attributes describe the entities
• What data types can be used to model the attributes
– Predefined – text, image, audio, video
Entity Relationship Diagram
• Technique for identifying entity types and
diagramming relationships between entity
types
• Determine predicates
– Teacher has office
– Professor teaches sections
– Course has sections
Data Views
• Logical view
– Soft view
• Screen output – query, report, etc.
– Hard view
• Printer output
• Physical view
– How data are stored on disk and what computer
has to do to find and retrieve data
Specialized Databases
• Multidimensional
– Single file
• Geographical Information System GIS
• Types
– Push or Pull
– Preprogrammed or ad hoc
Data Dictionaries
Meta Data
•
•
•
•
•
•
•
•
File Name
Field Name
Field Definition
Field Type
Column Heading
Field Length
Constraints
Validity Rules
Backup and Recovery
• Methods to Backup
– Disk or tape; on-site or off-site; electronic
• Disaster Recovery
– Cold sites and Hot sites
– Disaster Recovery Plans
Search Engine
• An insult to relational database to use term
to refer to search Engines
• A good topic for a future class – why search
engines fails the relational database test
Information Characteristics
• Quality
– Accuracy and precision; free of bias; complete,
timeliness, reliability of source
• Accessibility
– Availability, admissibility
• Presentation
– Level of summarization format
• Value and security