Indexing and Data Mining in Multimedia Databases

Table of Contents

Indexing and Data Mining in Multimedia Databases

Outline

Problem

Sample queries

Sample queries –cont’d

Outline

Indexing - Multimedia

PPT Slide

‘GEMINI’ - Pictorially

Remaining issues

Outline

FastMap

FastMap

Applications: time sequences

Applications - financial

Applications - financial

Application: VideoTrails

VideoTrails - usage

Application: VideoTrails

Outline

Merging similarity scores

‘FALCON’

“Single query point” methods

“Single query point” methods

PPT Slide

Conclusions for indexing + visualization

Outline

Data mining & fractals – Road map

Problem #1 - spatial d.m.

Problem#2: dim. reduction

Answer:

What is a fractal?

Definitions (cont’d)

Intrinsic (‘fractal’) dimension

Intrinsic (‘fractal’) dimension

Intrinsic (‘fractal’) dimension

Sierpinsky triangle

Road map

Solution#1: spatial d.m.

Solution#1: spatial d.m.

Solution#1: spatial d.m.

spatial d.m.

Solution#1: spatial d.m.

Solution#1: spatial d.m.

Problem #2: Dim. reduction

Solution:

Problem #2: dim. reduction

Problem #2: dim. reduction

Problem #2: dim. reduction

Road map

App. : traffic

More apps: Brain scans

More fractals:

More power laws

More power laws

Even more power laws:

Overall Conclusions:

Conclusions - cont’d

Related projects at CMU

Resources:

Email: informedia@cs.cmu.edu

Home Page: http://www.informedia.cs.cmu.edu

Other information:
Please send us email if you would like to request higher resolution slides.