Copyright © 2016 Ramez Elmasri and Shamkant B.

Navathe
 CHAPTER 26

Enhanced Data Models:

Introduction to Active,
Temporal, Spatial, Multimedia,
and Deductive Databases
Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe
26.1 Active Database Concepts and
Triggers
 Database systems implement rules that specify
actions automatically triggered by certain events
 Triggers
 Technique for specifying certain types of active
rules
 Commercial relational DBMSs have various
versions of triggers available
 Oracle syntax used to illustrate concepts

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 3

Generalized Model for Active
Databases and Oracle Triggers
 Event-condition-action (ECA) model
 Event triggers a rule
 Usually database update operations
 Condition determines whether rule action should
be completed
 Optional
 Action will complete only if condition evaluates to
true
 Action to be taken
 Sequence of SQL statements, transaction, or
external program
Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 4
Example
 Events that may cause a change in value of
Total_sal attribute
 Inserting new employee
 Changing salary
 Reassigning or deleting employees

Figure 26.1 A simplified COMPANY database used for active rule examples

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26-5

Example (cont’d.)
 Condition to be evaluated
 Check that value of Dno attribute is not NULL
 Action to be taken
 Automatically update the value of Total_sal

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26-6

Figure 26.2 Specifying active rules as triggers in Oracle notation (a) Triggers for
automatically maintaining the consistency of Total_sal of DEPARTMENT

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26-7

 Figure 26.2 (cont’d.) Specifying active rules as triggers in Oracle notation (b)
Trigger for comparing an employee’s salary with that of his or her supervisor

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26-8

Design and Implementation Issues
for Active Databases
 Deactivated rule
 Will not be triggered by the triggering event
 Activate command
 Makes the rule active again
 Drop command
 Deletes the rule from the system
 Approach: group rules into rule sets
 Entire rule set can be activated, deactivated, or
dropped

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 9

Design and Implementation Issues
for Active Databases (cont’d.)
 Timing of action
 Before trigger executes trigger before executing
event that caused the trigger
 After trigger executes trigger after executing the
event
 Instead of trigger executes trigger instead of
executing the event
 Action can be considered separate transaction
 Or part of same transaction that triggered the rule

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 10

Design and Implementation Issues
for Active Databases (cont’d.)
 Rule consideration
 Immediate consideration
 Condition evaluated as part of same transaction
 Evaluate condition either before, after, or instead of
executing the triggering event
 Deferred consideration
 Condition evaluated at the end of the transaction
 Detached consideration
 Condition evaluated as a separate transaction

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 11

Design and Implementation Issues
for Active Databases (cont’d.)
 Row-level rule
 Rule considered separately for each row
 Statement-level rule
 Rule considered once for entire statement
 Difficult to guarantee consistency and termination
of rules

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 12

Examples of Statement-Level Active
Rules in STARBURST

Figure 26.5 (continues) Active rules using statement-level

semantics in STARBURST notation
Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 13
Examples of Statement-Level Active
Rules in STARBURST (cont’d.)

Figure 26.5 (cont’d.) Active rules using statement-level

semantics in STARBURST notation

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 14

Potential Applications for Active
Databases
 Allow notification of certain conditions that occur
 Enforce integrity constraints
 Automatically maintain derived data
 Maintain consistency of materialized views
 Enable consistency of replicated tables

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 15

Triggers in SQL-99

Figure 26.6 Trigger T1 illustrating the syntax for defining triggers in SQL-99

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 16

26.2 Temporal Database Concepts
 Temporal databases require some aspect of time
when organizing information
 Healthcare
 Insurance
 Reservation systems
 Scientific databases
 Time considered as ordered sequence of points
 Granularity determined by the application

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 17

Temporal Database Concepts
(cont’d.)
 Chronon
 Term used to describe minimal granularity of a
particular application
 Reference point for measuring specific time
events
 Various calendars
 SQL2 temporal data types
 DATE, TIME, TIMESTAMP, INTERVAL, PERIOD

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 18

Temporal Database Concepts
(cont’d.)
 Point events or facts
 Typically associated with a single time point
 Time series data
 Duration events or facts
 Associated with specific time period
 Time period represented by start and end points
 Valid time
 True in the real world

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 19

Temporal Database Concepts
(cont’d.)
 Transaction time
 Value of the system clock when information is
valid in the system
 User-defined time
 Bitemporal database
 Uses valid time and transaction time
 Valid time relations
 Used to represent history of changes

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 20

Temporal Database Concepts
(cont’d.)

Figure 26.7 Different types of temporal relational databases (a) Valid time database
schema (b) Transaction time database schema (c) Bitemporal database schema

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 21

Temporal Database Concepts
(cont’d.)

Figure 26.8 Some tuple versions in the valid time relations EMP_VT and DEPT_VT

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 22

Temporal Database Concepts
(cont’d.)
 Types of updates
 Proactive
 Retroactive
 Simultaneous
 Timestamp recorded whenever change is applied
to database
 Bitemporal relations
 Application requires both valid time and
transaction time

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 23

Temporal Database Concepts
(cont’d.)
 Implementation considerations
 Store all tuples in the same table
 Create two tables: one for currently valid
information and one for the rest
 Vertically partition temporal relation attributes into
separate relations
 New tuple created whenever any attribute updated
 Append-only database
 Keeps complete record of changes and
corrections

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 24

Temporal Database Concepts
(cont’d.)
 Attribute versioning
 Simple complex object used to store all temporal
changes of the object
 Time-varying attribute
 Values versioned over time by adding temporal
periods to the attribute
 Non-time-varying attribute
 Values do not change over time

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 25

 Figure 26.10 Possible ODL schema for a temporal valid time EMPLOYEE_VT
object class using attribute versioning
Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 26
Temporal Database Concepts
(cont’d.)
 TSQL2 language
 Extends SQL for querying valid time and
transaction time tables
 Used to specify whether a relation is temporal or
nontemporal
 Temporal database query conditions may involve
time and attributes
 Pure time condition involves only time
 Attribute and time conditions

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 27

Temporal Database Concepts
(cont’d.)
 CREATE TABLE statement
 Extended with optional AS clause
 Allows users to declare different temporal options
 Examples:
 AS VALID STATE<GRANULARITY> (valid time
relation with valid time period)
 AS TRANSACTION (transaction time relation with
transaction time period)
 Keywords STATE and EVENT
 Specify whether a time period or point is
associated with valid time dimension
Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 28
Temporal Database Concepts
(cont’d.)
 Time series data
 Often used in financial, sales, and economics
applications
 Special type of valid event data
 Event’s time points predetermined according to
fixed calendar
 Managed using specialized time series
management systems
 Supported by some commercial DBMS packages

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 29

26.3 Spatial Database Concepts
 Spatial databases support information about
objects in multidimensional space
 Examples: cartographic databases, geographic
information systems, weather information
databases
 Spatial relationships among the objects are
important
 Optimized to query data such as points, lines,
and polygons
 Spatial queries

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 30

Spatial Database Concepts (cont’d.)
 Measurement operations
 Used to measure global properties of single
objects
 Spatial analysis operations
 Uncover spatial relationships within and among
mapped data layers
 Flow analysis operations
 Help determine shortest path between two points

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 31

Spatial Database Concepts (cont’d.)
 Location analysis
 Determine whether given set of points and lines lie
within a given polygon
 Digital terrain analysis
 Used to build three-dimensional models

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 32

Spatial Database Concepts (cont’d.)

Table 26.1 Common types of analysis for spatial data

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26-33

Spatial Database Concepts (cont’d.)
 Spatial data types
 Map data
 Geographic or spatial features of objects in a map
 Attribute data
 Descriptive data associated with map features
 Image data
 Satellite images
 Models of spatial information
 Field models
 Object models
Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 34
Spatial Database Concepts (cont’d.)
 Spatial operator categories
 Topological operators
 Properties do not change when topological
transformations applied
 Projective operators
 Express concavity/convexity of objects
 Metric operators
 Specifically describe object’s geometry
 Dynamic spatial operators
 Create, destroy, and update

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 35

Spatial Database Concepts (cont’d.)
 Spatial queries
 Range queries
 Example: find all hospitals with the Metropolitan
Atlanta city area
 Nearest neighbor queries
 Example: find police car nearest location of a crime
 Spatial joins or overlays
 Example: find all homes within two miles of a lake

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 36

Spatial Database Concepts (cont’d.)
 Spatial data indexing
 Grid files
 R-trees
 Spatial join index
 Spatial data mining techniques
 Spatial classification
 Spatial association
 Spatial clustering

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 37

26.4 Multimedia Database Concepts
 Multimedia databases allow users to store and
query images, video, audio, and documents
 Content-based retrieval
 Automatic analysis
 Manual identification
 Color often used in content-based image retrieval
 Texture and shape
 Object recognition
 Scale-invariant feature transform (SIFT) approach

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 38

Multimedia Database Concepts
(cont’d.)
 Semantic tagging of images
 User-supplied tags
 Automated generation of image tags
 Web Ontology Language (OWL) provides concept
hierarchy
 Analysis of audio data sources
 Text-based indexing
 Content-based indexing

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 39

26.5 Introduction to Deductive
Databases
 Deductive database uses facts and rules
 Inference engine can deduce new facts using rules
 Prolog/Datalog notation
 Based on providing predicates with unique names
 Predicate has an implicit meaning and a fixed
number of arguments
 If arguments are all constant values, predicate
states that a certain fact is true
 If arguments are variables, considered as a query or
part of a rule or constraint

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 40

Prolog Notation and The Supervisory
Tree

Figure 26.11 (a) Prolog notation (b) The supervisory tree

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26-41

Introduction to Deductive Databases
(cont’d.)
 Datalog notation
 Program built from basic objects called atomic
formulas
 Literals of the form p(a1,a2,…an)
 p is the predicate name
 n is the number of arguments for predicate p
 Interpretations of rules
 Proof-theoretic versus model-theoretic
 Deductive axioms
 Ground axioms
Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 42
Introduction to Deductive Databases
(cont’d.)

Figure 26.12 Proving a new fact

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26-43

Introduction to Deductive Databases
(cont’d.)
 Safe program or rule
 Generates a finite set of facts
 Nonrecursive query
 Includes only nonrecursive predicates

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 44

Use of Relational Operations

Figure 26.16
Predicates for
illustrating
relational operations

Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26-45

26.6 Summary
 Active databases
 Specify active rules
 Temporal databases
 Involve time concepts
 Spatial databases
 Involve spatial characteristics
 Multimedia databases
 Store images, audio, video, documents, and more
 Deductive databases
 Prolog and Datalog notation
Copyright © 2016 Ramez Elmasri and Shamkant B. Navathe Slide 26- 46