Transcript Lecture #15 - The University of Texas at Dallas

Data and Applications Security
Developments and Directions
Dr. Bhavani Thuraisingham
The University of Texas at Dallas
Lecture #15
Secure Multimedia Data Management and Data Mining
March 13, 2006
Objective
 This unit provides an overview of multimedia information
management including multimedia data management and multimedia
data mining. Security issues will also be discussed
 Reference: Managing and Mining Multimedia Databases, CRC Press,
Thuraisingham, June 2001
Outline
 Multimedia Data Management Systems
- Architecture
- Modeling
- Functions
 Security
 Developments and Challenges
 Multimedia Mining
 Future Directions
Sources of Multimedia Data
Text, Video, Audio, Imagery
Why Multimedia Database Management System?
 Need persistent storage for managing large quantities of multimedia
data
 A Multimedia DBMS manages multimedia data such as text, images,
audio, animation, video
 Extended by a Browser to produce a Hypermedia DBMS
 Heterogeneity with respect to data types
 Numerous Applications
- Entertainment, Defense and Intelligence, Telecommunications,
Finance, Medical
Architectures:
Loose Integration
User Interface
Module for Integrating
Data Manager with File Manager
Data Manager
for Metadata
Metadata
Multimedia
File Manager
Multimedia
Files
Architectures:
Tight Integration
User Interface
MM-DBMS:
Integrated data
manager and
file manager
Multimedia
Database
Architectures:
Functional
User Interface
• Representation
•
•
•
•
Distribution
Quality of Service
Real-time
Heterogeneity
Storage
• Query/Update
• Transactions
• Metadata
• Integrity/Security
Multimedia
Database
Data Model:
Scenario
Example:
Object A
2000 Frames
Object
representation
4/95
8/95
5/95
Object B
3000 Frames
10/95
Data Model:
Object
ID 2098
Object
A
interval (4/95, 8/95)
contents
Frames
2000
Data Model:
Object-Relational
ID
2098
Interval
(4/95, 8/95)
Contents
Frame
2000
Functions:
Editing
Example: Object editing
Editing objects A and B by merging them to form a new
object over interval 4/15/95 to 8/15/95
4/15/95
Object
C
8/15/95
Multimedia Data Access: Some approaches
 Text data
- Selection with index features
- Methods: Full text scanning, Inverted files, Document clustering
 Audio/Speech data
- Pattern matching algorithms

Matching index features given for searching and ones
available in the database
 Image data
- Identifying geometric boundaries, Identifying spatial
relationships, Image clustering
 Video data
- Retrieval with metadata, Pattern matching with images
Metadata for Multimedia
 Metadata may be annotations and stored in relations
- I.e., Metadata from text, images, audio and video are extracted
as stored as text
- Text metadata may be converted to relations by tagging and
extracting concepts
 Metadata may be images of video data
- E.g., certain frames may be captured as metadata
 Multimedia data understanding
- Extracting metadata from the multimedia data
Storage Methods
 Single disk storage
- Objects belonging to different media types in same disk
 Multiple disk storage
- Objects distributed across disks

Example: individual media types stored in different disks

I.e., audio in one disk and video in another

Need to synchronize for presentation (real-time techniques)
 Multiple disks with striping
- Distribute placement of media objects in different disks

Called disk striping
Security Issues
 Access Control
 Multilevel Security
 Architecture
 Secure Geospatial Information Systems
Access Control for Multimedia Databases
 Access Control for Text, Images, Audio and Video
 Granularity of Protection
- Text

John has access to Chapters 1 and 2 but not to 3 and 4
- Images

John has access to portions of the image

Access control for pixels?
- Video and Audio

John has access to Frames 1000 to 2000

Jane has access only to scenes in US
- Security constraints

Association based constraints
E.g., collections of images are classified
MLS Security
Book
Object
References
Introduction
Set of Sections
Introduction: Level = Unclassified
Set of Sections: Level = TopSecret
References: Level = Secret
Example Security Architecture: Integrity Lock
Trusted Agent
to compute
checksums
Untrusted
Sensor
Multimedia
Data
Data Manager
Manager
Compute Checksum
Based on say multimedia data value
(such as video object content)
Security level and Checksum
Multimedia
Database
Compute Checksum
Based on multimedia data value
and Security level retrieved
from the stored multimedia database
Inference Control
User Interface Manager
Metadata,
Constraints
Multimedia
Database
Inference Engine
Acts as an Inference
Controller
Multimedia
Database
Manager
Authorization Model for Secure Geospatial
Systems
 Geospatial images could be Digital Raster Images that store images
as pixels or Digital Vector Images that store images as points, lines
and polygons
 GSAM: Geospatial Authorization Model specifies subjects,
credentials, objects (e.g, points, lines, pixels etc.) and the access
that subjects have to objects
 Reference: Authorization Model for Geospatial Data; Atluri and Chun,
IEEE Transactions on Dependable and Secure Computing, Volume 1,
#4, October – December 2004.
Secure Geospatial Systems
++++
++++
++++
++++
++++
++++
Classified content blanked at the Unclassified level
++++
Unclassified content
Directions and Challenges in Managing
Multimedia Databases
 Much work on data models, query languages, architectures and
indexing (still need more work on indexing)
 Increasing interest in
- Quality of Service for Multimedia Data Management

Synchronizing audio and video

Synchronizing storage retrieval and presentations

Real-time scheduling techniques
- Distributed multimedia database management

Query processing techniques
- Multimedia on the Web

Capture, annotate, summarize, disseminate
- Mining multimedia data

Extracting information previously unknown
Example: Automated Digital Capture, Analysis
and Publication of Broadcast News
Video
Source
Broadcast News Editor (BNE)
Scene
Change
Detection
Frame
Classifier
Imagery
Silence
Detection
Correlation
Story
GIST Theme
Broadcast
Detection
Commercial
Detection
Key Frame
Selection
Story
Segmentation
Audio
Closed
Caption
Text
Speaker
Change
Detection
Closed
Caption
Preprocess
Segregate
Video
Streams
Broadcast News
Navigator (BNN)
Token
Detection
Named
Entity
Tagging
Analyze and Store Video and Metadata
Multimedia
Database
Management
System
Video
and
Metadata
Web-based Search/Browse by
Program, Person, Location, ...
Example Web Page
Select
Story
Elaborate on Story
Source
Key
Frame
Closed
Caption
6 Most
Frequent
Tags
Related Web Sites
Video
Summary
Apply QueryFlocks Data Mining Tool:
(MITRE/Stanford Tool)
Extracting Relations from Text for Mining:
An Example
Text
Corpus
Concept
Extraction
Goal: Find
Cooperating/
Combating Leaders
in a territory
Association
Rule
Product
Repository
Person1
Natalie Allen
Leon Harris
Ron Goldman
Mobotu Sese
Seko
Person2
Linden Soles
Joie Chen
Nicole Simpson
...
Laurent Kabila
117
53
19
10
Image Processing:
Example: Change Detection:
 Trained Neural Network to predict “new” pixel from “old” pixel
- Neural Networks good for multidimensional continuous data
- Multiple nets gives range of “expected values”
 Identified pixels where actual value substantially outside range of
expected values
- Anomaly if three or more bands (of seven) out of range
 Identified groups of anomalous pixels
In Conclusion:
 Multimedia data management is getting mature
 Numerous applications in several industries
 Challenge is to mine multimedia databases
 Work is just beginning on multimedia data mining
 Web provides lots of opportunities and challenges for
multimedia data management
 We cannot forget about security and privacy