A primer on traditional hard and emerging soft computing approaches for mining multimedia data

While the digital revolution has made huge volumes of high dimensional multimedia data available, it has also challenged users to extract the information they seek from heretofore unthinkably huge datasets. Traditional hard computing data mining techniques have concentrated on flat-file applications. Soft computing tools–such as fuzzy sets, artificial neural networks, genetic algorithms, and rough sets–however, offer the opportunity to apply a wide range of data types to a variety of vital functions by handling real-life uncertainty with low-cost solutions. Data Mining: Multimedia, Soft Computing, and Bioinformatics provides an accessible introduction to fundamental and advanced data mining technologies.

This readable survey describes data mining strategies for a slew of data types, including numeric and alpha-numeric formats, text, images, video, graphics, and the mixed representations therein. Along with traditional concepts and functions of data mining–like classification, clustering, and rule mining–the authors highlight topical issues in multimedia applications and bioinformatics.
Principal topics discussed throughout the text include:

• The role of soft computing and its principles in data mining
• Principles and classical algorithms on string matching and their role in data (mainly text) mining
• Data compression principles for both lossless and lossy techniques, including their scope in data mining
• Access of data using matching pursuits both in raw and compressed data domains
• Application in mining biological databases

Contents

• Introduction to Data Mining
• Soft Computing
• Multimedia Data Compression
• String Matching
• Classification in Data Mining
• Clustering in Data Mining
• Association Rules
• Rule Mining with Soft Computing
• Multimedia Data Mining
• Bioinformatics: An Application
• Index
• About the Authors