Managing Data Resources Chapter Seven. SoftwareInformation Systems for Management2 Hierarchy of Data Bit Byte Field Record File Database Database management.

Managing Data Resources

Chapter Seven

Software Information Systems for Management

2

Hierarchy of Data

• Bit• Byte• Field• Record• File• Database• Database management system


3

Traditional Data Environment

• Files for each application/department/function• Duplicated/redundant data/files• Inability to link data/files• Program-data dependence (change one, you

must change the other)


4

Problems with Traditional Data Environment

Data redundancy leads to• Lack of data integrity• Program-data dependence• Lack of flexibility (no ad hoc reports/different

views• Poor security (access)• Lack of data sharing & availability


5

DBMS Approach

• Database: Collection of one or more files containing data organized to serve multiple applications by minimizing redundant data.

• Database management system controls organization of & access to data and database files by acting as interface between the data & application programs and as an environment for developing and using databases.


6

Views/Schemas

• Logical view: How end users perceive the data is organized– Schema: The view of all the data– Subschema: A partial view of the data accessible to

an end user (e.g., “view only” a subset of screens/data)

• Physical view: How the data are actually organized on physical storage media


7

Components of a DBMS #1

• Data definition language (DDL)– Formal language associated with DBMS– Used by both end users & programmers to

manipulate data

• Data manipulation language– Commands to modify/extract data & to develop apps– Structured Query Language (SQL)– Can use various languages in addition to/ instead of

SQL


8

Components of a DBMS #2

• Data dictionary

–Defines each data element (# bytes, text/numeric, etc., format, range, access, use, ownership, physical representation)

–Used for communication between developers & users and for standardization of data/databases/ programs

–Some data dictionaries are active; changes automatically change related databases/programs

NOTE: Any properly developed information system should have a data dictionary.


9

Database Models (Types) #1

– Hierarchical• Upside-down tree-like structure• Root (top most data element) is the key field• Each child record can have only one parent record (1:M relationships); parents can have many children• Pointers for expressing relationships• Hard to change & limited retrieval capabilities• “Legacy” systems

– Network• Similar to hierarchical but M:M relationships between

records• Complex and hard to change


10

Database Models (Types) #2

• Relational– 2-dimensional tables (relations)– Physically appear similar to files (but are not)– Row/record/tuple– Column/field/attribute/data element– Ability to link relations on-the-fly

• Select creates a subset of all records that meet specified criteria

• Join combines tables into a single new table• Project creates a subset of columns in a table, resulting in

new tables/views


11

SQL

• Principal data manipulation language for relational DBMS

• Versions that can run on almost any OS & computer (mainframe, PC, etc.)

• Easy to learn & use– Select lists desired columns from desired table(s)– From identifies tables/views from which to select

columns– Where are conditions for selecting specific records &

for joining multiple tables


12

SQL Example

SELECT Part:Part_Number, Supplier:Supplier_Number, Supplier:Supplier_Name, Supplier:Supplier_Address

FROM Part, Supplier

WHERE Part:Supplier_Number=Supplier:Supplier_Number AND Part_Number=137 OR Part_Number=152

Note: No line returns in any of the commands


13

Object-Oriented Databases (OODBMS)

• Store data & procedures that act on the data as objects that can be automatically retrieved & shared

• Objects can contain multimedia• Object-relational databases: Relational

databases that can store both traditional data & object-oriented data that store graphics & multimedia


14

Designing Databases

• Entity-relationship diagram (E-R diagram)– Documents database by showing relationships

among entities in database

• Normalization– Creates small, stable data structures (tables) from

complex groups of data– Example: Student data: Normalization results in

several DB tables of student data: Name/address, Courses taken, Funds/received/distributed, etc.

– See Figures 7-14 & 7-15

E-R Diagram

ORDER

PART

Includes

Delivered by

SUPPLIER

1

M

M

1Figure 7-13, p. 217


16

Distributed Databases

• Stored in more than one physical location• Reduce vulnerability• Increase responsiveness• Can run on cheaper computer systems• Weakness: Vulnerability of telecommunications• Sometimes, locals can depart from acceptable DB

practices– Partitioned: Each remote processor has its own necessary data– Duplicated: Duplicated database (reconciled periodically)


17

Data Administration

• Information policy: Planning & rules governing DB operations & information use

• Data planning: Enterprise analysis• Maintenance of data dictionaries• Data quality standards• Database Administration

– Technical– Operational– May include personnel, purchasing, etc. for DB

function


18

DB Trends

• Multidimensional Data Analysis– Online analytical processing (OLAP)

• Ability to “slice and dice” data interactively• Multiple perspectives• Matrices or cubes


19

Data Warehouses & Datamining

• Data warehouse– Consolidates & stores current & historical data

extracted from various operational systems– Meta-data (summaries of transactional data)– Reporting & query tools, including OLAP & data

mining

• Data mart: Subset of data warehouse• Datamining: Analysis of data in data

warehouses to find patterns/rules to aid in decision making


20

Databases & the Web

• Hypermedia database– Organize data in network of nodes linked in user-

specified patterns/relationships– Text, graphics, sound, video, programs

• Linking internal DBs to the Web– Middleware is interface between DB & browser– Application server uses middleware to interface

between DB & browser– Common gateway interface (CGI) written in a

language interfaces between DB, app server, & browser


21

Ford & Firestone: Tire Disaster

• Late September, 2001: Firestone recalled another 3.5 million Wilderness tires

• How does this crisis represent an information management problem?

• Why did management of these two companies (and government officials) not see the trends in the data?

• What would you suggest should have been done in terms of database management & queries?


22

Next Classes:

• October 17: Midterm– Bring a pencil for Scantron sheets– Bring a pen if you prefer to answer short answer

questions in pen rather than pencil– Room 343: 6-8:00 p.m.– Bring your UM ID; it will be checked

• October 22– Chapter 8: Telecommunications & Networks– Case Study: Monitoring Employees on Networks

Managing Data Resources Chapter Seven. SoftwareInformation Systems for Management2 Hierarchy of Data Bit Byte Field Record File Database Database management.

Documents

data dictionary

data dictionaries

data accessible

data subschema

data application programs

softwareinformation

management2 hierarchy

data resources chapter