Preface
The digital revolution
and the explosive growth of the Internet have helped create collection of huge
amounts of useful data of diverse characteristics. Data is a valuable and
intangible asset in any business today. Databases and database technologies
play a crucial role in maintaining and manipulating data. Various database
models exist such as relational, hierarchical, network, flat, and
object-oriented. Each model organizes data in a different way to make them suitable
for the intended application.
Real
world data are diverse and imprecise in nature. They are growing at a
phenomenal rate. As the application needs are diverse they demand a completely
different set of requirements on the underlying models. The conventional
relational database model is no longer appropriate for heterogeneous data. The
diverse characteristics of data and its huge volume demand new ways of carrying
out data analysis. Soft computing is a new, emerging complementary discipline
for traditional computing principles. It exploits the tolerance for imprecision
and uncertainty to achieve solutions for complex problems.
Soft
computing methodologies include fuzzy sets, neural networks, genetic
algorithms, Bayesian belief networks and rough sets. Fuzzy sets provide a
natural framework for dealing with uncertainty. Bayesian belief networks,
neural networks, and rough sets are widely used for classification and rule
generation. Genetic algorithms handle various optimization and search processes
like query optimization and template selection. Rough sets handle uncertainty
arising from granules in the domain of discourse.
The
advent of soft computing marks a significant paradigm shift in computing.
Currently it has a wide range of application. The various techniques of soft
computing and their applications in database technologies are discussed in the
chapters of this book.
The
first chapter, titled “Fuzzy Database Modeling: An Overview and New
Definitions,” by Angélica Urrutia and José Galindo gives an overview of fuzzy
database modeling. Much of data in real world is not precise, but fuzzy.
Zadeh’s fuzzy logic gives a tool to handle fuzzy data in decision making.
Modeling of fuzzy data has been studied by a number of researchers. The authors
of this chapter discuss further extensions in the field of fuzzy database
modelling. This chapter starts with a review of contributions of previous
authors with respect to fuzzy database modeling, particularly the FuzzyEER
model. The chapter then proceeds to introduce and explain their proposed newly
definitions pertaining to the FuzzyEER Model regarding fuzzy attributes, fuzzy
degrees, and fuzzy entities. The newly introduced concepts are amply
illustrated through suitable examples. The authors hope that the new definitions
will further enhance FuzzyEER model to facilitate fuzzy queries and fuzzy data
mining.
The
author Pierre Collet, in the chapter “A quick presentation of Evolutionary
Computation,” gives an easy-to-grasp exposition of generic evolutionary
computation paradigm. After giving a brief historical perspective, the author
presents a unified evolutionary algorithm. The various concepts involved are
lucidly explained with examples. This chapter will be quite useful to get a
gentle introduction and survey of generic evolutionary computation.
The
third chapter, “Evolutionary Algorithms in Supervision of Error-free Control,”
by Bohumil Sulc and David Klimanek, reports application of certain soft
computing techniques in combustion control. Specifically, genetic and simulated
annealing algorithms have been employed in a model-based controlled variable
sensor discredibility detection. The authors outline procedure of incorporating
genetic and simulated annealing algorithms in the control loop. They claim that
such application of these soft computing techniques has a great importance in
industrial practice because a timely predicted sensor malfunction helps to save
additional costs resulting from unplanned shutdowns.
In
the next chapter titled “Soft Computing Techniques in Spatial Databases,” its
author Markus Schneider explains how two different soft computing techniques
with different expressiveness can be used for spatial data handling in the
context of spatial databases and Geographic Information Systems. The focus of this
chapter is the design of the algebra systems named Vague Spatial Algebra (VASA)
and Fuzzy Spatial Algebra (FUSA). A formal definition of the structure and
semantics of both types of systems are also provided. Further, spatial set
operations for both the algebras have also been discussed. Finally, a
description of how these data types can be embedded into extensible databases
is explained with sample queries.
In
the fifth chapter, “Type-2 Fuzzy Interface for Artificial Neural Network”, the
author Priti Srinivas Sajja introduces another hybrid soft computing technique.
The field of applications of soft computing technique discussed by the author
is the process of course selection performed by students. The author introduces
a generic framework of type-2 fuzzy interface to an Artificial Neural Network
system. The author covers the introduction of fuzzy logic, fuzzy membership
functions, type-1 and type-2 fuzzy systems and Artificial Neural Networks (ANN)
for novice readers. Also, the author gives the need for hybridization of ANN
and fuzzy logic. Next, the author illustrates an experimental prototype with
fuzzy interface and base ANN. In this system, the author uses type-2 fuzzy
interface to feed input to the base ANN. The author claims that with sufficient
amount of good quality input data, the system performs well and with minor
modification, the system may be used for HR Management, aptitude testing and
general career counseling.
In
chapter VI, “A Combined GA-Fuzzy Classification System for Mining Gene Expression
Databases,” the authors Gerald Schaefer and Tomoharu Nakashima introduce yet
another hybrid soft computing technique. The field of application of soft
computing technique discussed by the authors is gene expression database. After
explaining fuzzy rule generation and fuzzy rule classification, the authors
point out the very many fuzzy if-then rules that would result. The number of
generated rules increases exponentially with the number of attributes involved
and with the number of partitions used for each attribute. The genetic
algorithm technique is employed to reduce the number of rules to a compact set.
The authors discuss genetic operations employed and give their algorithm in
detail and ways of improving its performance. The authors demonstrate their
hybrid technique on three gene expression data sets that are commonly used in
the literature, viz., Colon dataset, Leukemia dataset and Lymphoma dataset.
Exhaustive simulation results are given. The authors state that their technique
yields good results.
The
chapter on “Fuzzy Decision Rule Construction Using Fuzzy Decision Trees (FDT):
Application to E-Learning Database” by Malcolm J. Beynon and Paul Jones
(Chapter VII) discusses an application of fuzzy logic–specifically fuzzy
decision tree. The authors give two sets of extensive FDT analysis. The first
analysis deals with a small example dataset to illustrate the concepts. The
second FDT analysis is in the field of E-Learning and considers the student’s
weekly online activities and subsequent performance in a university course. The
authors emphasize visualization of results throughout the chapter.
Chapter
VIII, “A Bayesian Belief Network Methodology for Modeling Social Systems in
Virtual Communities: Opportunities for Database Technologies,” is contributed
by the authors Ben K. Daniel, Juan-Diego Zapata-Rivera, and Gordon I.Mc Calla.
Bayesian belief networks are used to model situations involving uncertainty
arising in fields like social sciences. A Bayesian model encodes domain
knowledge, showing relationships, interdependencies, and independencies among
variables. However, knowledge engineering effort is required to create
conditional probability for each variable in the network. This chapter
describes an approach that combines both qualitative and quantitative
techniques to elicit knowledge from experts without worrying about computing
initial probabilities for training the model. The authors demonstrate their
technique on a computational model of social capital in virtual communities.
The
importance of preserving the accuracy and integrity of data in a database is
highlighted in the chapter titled “Integrity Constraints Checking in a
Distributed Database” by Hamidah Ibrahim. This chapter discusses checking
integrity constraints in distributed databases. The author describes different
integrity constraint tests that could be conducted in distributed databases. A
review of the integrity constraints available in the literature is clearly
given. Finally the author explains several strategies for checking the
integrity constraints in distributed databases. The author also discusses
important criteria for evaluating the integrity tests.
Authors
G. Castellano, A. M. Fanelli, and M. A. Torsello in chapter X, “Soft Computing
techniques in Content-Based Multimedia Information Retrieval,” clearly explain
the basic concepts of the four techniques coming under the purview of
softcomputing, viz., Fuzzy Logic, Neural Networks, Rough Sets, and Genetic
Algorithm. They give a good literature survey on the application of each soft
computing technique to Content Based Multimedia Information Retrieval(CB-MIR).
They also give a good survey of the applicaiton of hybrid, neuro-fuzzy
technique to CB-MIR. Next, the authors discuss in detail about applying hybrid
neuro-fuzzy techniques to CB-MIR. They have contributed a system called
VIRMA(Visual Image Retrieval by Shape MAtching) that enables users to search
for images having a shape similar to the sketch of a submitted sample image.
The neuro-fuzzy strategy enables one to extract a set of fuzzy rules that
classify image pixels for the extraction of contours included in the processed
image, so that this can be stored in the database. The authors show how the
neuro-fuzzy technique is useful for CB-MIR.
Chapter
XI, “An Exposition of Feature Selection and Variable Precision Rough Set
Analysis: Application to Financial Data,” by Malcolm J. Beynon and Benjamin
Griffiths presents a Variable Precision Rough Sets (VPRS) analysis of certain
Fitch Individual Bank Rating (FIBR) datasets. There are two parts of
elucidation undertaken in this chapter. First, the levels of possible
pre-processing necessary for undertaking a Rough Set based Theory (RST)
analysis and then the presentation of an analysis using VPRS. The vein graph
software enables one to select a single β-reduct and derive the rules
associated with the β-reduct. Two algorithms are used for feature selection
namely ReliefF and RST_FS. The predictions based on the training and validation
sets are displayed in the ‘Predictive Summary Stat’s panel.
Xenia
Nadenova introduces the concept of Good Diagnostic Test (GDT) as the basis of
her approach in “Interconnection of Class of Machine Learning Algorithms with
Logical Commonsense Reasoning Operations” in chapter XII. This chapter explains
the possibility of transforming a large class of machine learning algorithms
into a commonsense reasoning processes by using well-known deduction and
induction logical rules. The lattice theory is used for constructing a good
classification test. The rules for implementing variant transitions have been
constructed such as rules of generalization and specialization, inductive
diagnostic rules, and dual inductive diagnostic rules. The commonsense
reasoning rules have been divided into two classes. An algorithm DIAGaRa for
inferring GMRTs has been proposed for incremental inferring of good diagnostic
tests. The algorithm for inferring good tests is decomposed into subtasks and
operations that are in accordance with main human common sense reasoning rules.
In
chapter XIII, the authors Erhan Akdoğan, M. Arif Adlı, Ertuğrul Taçgın, and
Nureddin Bennett propose a human-machine interface (HMI) to control a robot
manipulator that has three-degrees of freedom for the rehabilitation of the
lower limbs. This system uses a rule-based intelligent controller structure,
combined with conventional control algorithms. It also has a user friendly GUI
which can be used on the Internet, thereby allowing the patients to receive
treatment at home. With HMI, the progress and the current state of a patient’s
rehabilitation can be stored in the database. The system proposed in this
chapter can handle common problems such as the transportation of patients,
storage of data and availability of data of the progress of patient’s
rehabilitation. The authors claim that by utilizing this system
physiotherapists can treat several patients at the same time.
In
the last chapter, “Congestion Control using Soft Computing”, the authors
T.Revathi, and K.Muneeswaran discuss the phenomenon of network congestion. They
recapitulate some of the important existing techniques for congestion control.
Then they take up the congestion avoidance problem and explain the need for
Active Queue Management (AQM). They review some of the AQM techniques available
in the literature. Then the authors propose a soft computing technique called
Fuzzy-enabled Active Queue Management (F-AQM) which addresses the influence of
the queuing behavior in handling the traffic in a network. They design a fuzzy
rule base represented in the form of a matrix indexed by queue length and rate
of change of queue. They have studied the performance of their scheme by
suitable simulation and compared the performance with that of Adaptive Virtual
Queue (AVQ) techniques. It is claimed that the proposed method outperforms AVQ
in reducing the number of dropped packets for different settings of Explicit
Congestion Notification (ECN) and queue size.
Author(s)/Editor(s)
Biography
K. Anbumani
Kalirajan Anbumani obtained his Bachelor of Engineering
degree from the University of Madras (1962), Master of Engineering degree from
the University of Pune (1967) and Ph.D degree from the Indian Institute of Science,
Bangalore – all in India. Initially, he served in the industry for 2 years.
Subsequently, he took up engineering teaching in government and private
engineering colleges, Bharathiar university and Karunya University. After
holding his last position as the Director, School of Computer Science and
Technology, Karunya University, he has recently taken time off to engage in
writing a book. Prof. Anbumani has many research publications, including
chapters in books, in areas such as information security, data mining, data
compression, multimedia information retrieval, soft-computing, object-oriented
methodology, real-time systems and control. He has completed many funded
projects and has conducted a number of conferences and chaired conference
sessions. Current interest of Prof. Anbumani covers security, including data
hiding in multimedia.
R. Nedunchezhian
Raju Nedunchezhian is
currently working as the Vice-Principal of Kalaignar Karunanidhi Institute of
Technology, Coimbatore, TamilNadu, India. Previously, he served as Research
Coordinator of the Institute and Head of Computer Science and Engineering
Department (PG) at Sri Ramakrishna Engineering College, Coimbatore. He has more
than 17 years of experience in research and teaching. He obtained his BE(Computer
Science and Engineering) degree in the year 1991, ME(Computer Science and
Engineering) degree in the year 1997 and Ph.D(Computer Science and Engineering)
in the year 2007. He has guided numerous UG, PG and M.Phil projects and
organized a few sponsored conferences and workshops funded by private and
government agencies. Currently, he is guiding many Ph.D scholars of the Anna
University, Coimbatore and the Bharathiar University. His research interests
include knowledge discovery and data mining, Soft Computing, distributed
computing and database security. He has published many research papers in
national/international conferences and journals. He is a Life member of
Advanced Computing and Communication Society and ISTE.
Reviews
and Testimonials
"The advent of
soft computing marks a significant paradigm shift in computing. Currently it
has a wide range of application. The various techniques of soft computing and
their applications in database technologies are discussed in the chapters of
this book."
- K.
Anbumani
Name : I Gede Alit Yoga Sedana
