\documentclass[11pt,twoside]{article}\makeatletter \IfFileExists{xcolor.sty}% {\RequirePackage{xcolor}}% {\RequirePackage{color}} \usepackage{colortbl} \usepackage{wrapfig} \usepackage{ifxetex} \ifxetex \usepackage{fontspec} \usepackage{xunicode} \catcode`⃥=\active \def⃥{\textbackslash} \catcode`❴=\active \def❴{\{} \catcode`❵=\active \def❵{\}} \def\textJapanese{\fontspec{Noto Sans CJK JP}} \def\textChinese{\fontspec{Noto Sans CJK SC}} \def\textKorean{\fontspec{Noto Sans CJK KR}} \setmonofont{DejaVu Sans Mono} \else \IfFileExists{utf8x.def}% {\usepackage[utf8x]{inputenc} \PrerenderUnicode{–} }% {\usepackage[utf8]{inputenc}} \usepackage[english]{babel} \usepackage[T1]{fontenc} \usepackage{float} \usepackage[]{ucs} \uc@dclc{8421}{default}{\textbackslash } \uc@dclc{10100}{default}{\{} \uc@dclc{10101}{default}{\}} \uc@dclc{8491}{default}{\AA{}} \uc@dclc{8239}{default}{\,} \uc@dclc{20154}{default}{ } \uc@dclc{10148}{default}{>} \def\textschwa{\rotatebox{-90}{e}} \def\textJapanese{} \def\textChinese{} \IfFileExists{tipa.sty}{\usepackage{tipa}}{} \fi \def\exampleFont{\ttfamily\small} \DeclareTextSymbol{\textpi}{OML}{25} \usepackage{relsize} \RequirePackage{array} \def\@testpach{\@chclass \ifnum \@lastchclass=6 \@ne \@chnum \@ne \else \ifnum \@lastchclass=7 5 \else \ifnum \@lastchclass=8 \tw@ \else \ifnum \@lastchclass=9 \thr@@ \else \z@ \ifnum \@lastchclass = 10 \else \edef\@nextchar{\expandafter\string\@nextchar}% \@chnum \if \@nextchar c\z@ \else \if \@nextchar l\@ne \else \if \@nextchar r\tw@ \else \z@ \@chclass \if\@nextchar |\@ne \else \if \@nextchar !6 \else \if \@nextchar @7 \else \if \@nextchar (8 \else \if \@nextchar )9 \else 10 \@chnum \if \@nextchar m\thr@@\else \if \@nextchar p4 \else \if \@nextchar b5 \else \z@ \@chclass \z@ \@preamerr \z@ \fi \fi \fi \fi \fi \fi \fi \fi \fi \fi \fi \fi \fi \fi \fi \fi} \gdef\arraybackslash{\let\\=\@arraycr} \def\@textsubscript#1{{\m@th\ensuremath{_{\mbox{\fontsize\sf@size\z@#1}}}}} \def\Panel#1#2#3#4{\multicolumn{#3}{){\columncolor{#2}}#4}{#1}} \def\abbr{} \def\corr{} \def\expan{} \def\gap{} \def\orig{} \def\reg{} \def\ref{} \def\sic{} \def\persName{}\def\name{} \def\placeName{} \def\orgName{} \def\textcal#1{{\fontspec{Lucida Calligraphy}#1}} \def\textgothic#1{{\fontspec{Lucida Blackletter}#1}} \def\textlarge#1{{\large #1}} \def\textoverbar#1{\ensuremath{\overline{#1}}} \def\textquoted#1{‘#1’} \def\textsmall#1{{\small #1}} \def\textsubscript#1{\@textsubscript{\selectfont#1}} \def\textxi{\ensuremath{\xi}} \def\titlem{\itshape} \newenvironment{biblfree}{}{\ifvmode\par\fi } \newenvironment{bibl}{}{} \newenvironment{byline}{\vskip6pt\itshape\fontsize{16pt}{18pt}\selectfont}{\par } \newenvironment{citbibl}{}{\ifvmode\par\fi } \newenvironment{docAuthor}{\ifvmode\vskip4pt\fontsize{16pt}{18pt}\selectfont\fi\itshape}{\ifvmode\par\fi } \newenvironment{docDate}{}{\ifvmode\par\fi } \newenvironment{docImprint}{\vskip 6pt}{\ifvmode\par\fi } \newenvironment{docTitle}{\vskip6pt\bfseries\fontsize{22pt}{25pt}\selectfont}{\par } \newenvironment{msHead}{\vskip 6pt}{\par} \newenvironment{msItem}{\vskip 6pt}{\par} \newenvironment{rubric}{}{} \newenvironment{titlePart}{}{\par } \newcolumntype{L}[1]{){\raggedright\arraybackslash}p{#1}} \newcolumntype{C}[1]{){\centering\arraybackslash}p{#1}} \newcolumntype{R}[1]{){\raggedleft\arraybackslash}p{#1}} \newcolumntype{P}[1]{){\arraybackslash}p{#1}} \newcolumntype{B}[1]{){\arraybackslash}b{#1}} \newcolumntype{M}[1]{){\arraybackslash}m{#1}} \definecolor{label}{gray}{0.75} \def\unusedattribute#1{\sout{\textcolor{label}{#1}}} \DeclareRobustCommand*{\xref}{\hyper@normalise\xref@} \def\xref@#1#2{\hyper@linkurl{#2}{#1}} \begingroup \catcode`\_=\active \gdef_#1{\ensuremath{\sb{\mathrm{#1}}}} \endgroup \mathcode`\_=\string"8000 \catcode`\_=12\relax \usepackage[a4paper,twoside,lmargin=1in,rmargin=1in,tmargin=1in,bmargin=1in,marginparwidth=0.75in]{geometry} \usepackage{framed} \definecolor{shadecolor}{gray}{0.95} \usepackage{longtable} \usepackage[normalem]{ulem} \usepackage{fancyvrb} \usepackage{fancyhdr} \usepackage{graphicx} \usepackage{marginnote} \renewcommand{\@cite}[1]{#1} \renewcommand*{\marginfont}{\itshape\footnotesize} \def\Gin@extensions{.pdf,.png,.jpg,.mps,.tif} \pagestyle{fancy} \usepackage[pdftitle={Encapsulation of Soft Computing Approaches within Itemset Mining -A Survey}, pdfauthor={}]{hyperref} \hyperbaseurl{} \paperwidth210mm \paperheight297mm \def\@pnumwidth{1.55em} \def\@tocrmarg {2.55em} \def\@dotsep{4.5} \setcounter{tocdepth}{3} \clubpenalty=8000 \emergencystretch 3em \hbadness=4000 \hyphenpenalty=400 \pretolerance=750 \tolerance=2000 \vbadness=4000 \widowpenalty=10000 \renewcommand\section{\@startsection {section}{1}{\z@}% {-1.75ex \@plus -0.5ex \@minus -.2ex}% {0.5ex \@plus .2ex}% {\reset@font\Large\bfseries}} \renewcommand\subsection{\@startsection{subsection}{2}{\z@}% {-1.75ex\@plus -0.5ex \@minus- .2ex}% {0.5ex \@plus .2ex}% {\reset@font\Large}} \renewcommand\subsubsection{\@startsection{subsubsection}{3}{\z@}% {-1.5ex\@plus -0.35ex \@minus -.2ex}% {0.5ex \@plus .2ex}% {\reset@font\large}} \renewcommand\paragraph{\@startsection{paragraph}{4}{\z@}% {-1ex \@plus-0.35ex \@minus -0.2ex}% {0.5ex \@plus .2ex}% {\reset@font\normalsize}} \renewcommand\subparagraph{\@startsection{subparagraph}{5}{\parindent}% {1.5ex \@plus1ex \@minus .2ex}% {-1em}% {\reset@font\normalsize\bfseries}} \def\l@section#1#2{\addpenalty{\@secpenalty} \addvspace{1.0em plus 1pt} \@tempdima 1.5em \begingroup \parindent \z@ \rightskip \@pnumwidth \parfillskip -\@pnumwidth \bfseries \leavevmode #1\hfil \hbox to\@pnumwidth{\hss #2}\par \endgroup} \def\l@subsection{\@dottedtocline{2}{1.5em}{2.3em}} \def\l@subsubsection{\@dottedtocline{3}{3.8em}{3.2em}} \def\l@paragraph{\@dottedtocline{4}{7.0em}{4.1em}} \def\l@subparagraph{\@dottedtocline{5}{10em}{5em}} \@ifundefined{c@section}{\newcounter{section}}{} \@ifundefined{c@chapter}{\newcounter{chapter}}{} \newif\if@mainmatter \@mainmattertrue \def\chaptername{Chapter} \def\frontmatter{% \pagenumbering{roman} \def\thechapter{\@roman\c@chapter} \def\theHchapter{\roman{chapter}} \def\thesection{\@roman\c@section} \def\theHsection{\roman{section}} \def\@chapapp{}% } \def\mainmatter{% \cleardoublepage \def\thechapter{\@arabic\c@chapter} \setcounter{chapter}{0} \setcounter{section}{0} \pagenumbering{arabic} \setcounter{secnumdepth}{6} \def\@chapapp{\chaptername}% \def\theHchapter{\arabic{chapter}} \def\thesection{\@arabic\c@section} \def\theHsection{\arabic{section}} } \def\backmatter{% \cleardoublepage \setcounter{chapter}{0} \setcounter{section}{0} \setcounter{secnumdepth}{2} \def\@chapapp{\appendixname}% \def\thechapter{\@Alph\c@chapter} \def\theHchapter{\Alph{chapter}} \appendix } \newenvironment{bibitemlist}[1]{% \list{\@biblabel{\@arabic\c@enumiv}}% {\settowidth\labelwidth{\@biblabel{#1}}% \leftmargin\labelwidth \advance\leftmargin\labelsep \@openbib@code \usecounter{enumiv}% \let\p@enumiv\@empty \renewcommand\theenumiv{\@arabic\c@enumiv}% }% \sloppy \clubpenalty4000 \@clubpenalty \clubpenalty \widowpenalty4000% \sfcode`\.\@m}% {\def\@noitemerr {\@latex@warning{Empty `bibitemlist' environment}}% \endlist} \def\tableofcontents{\section*{\contentsname}\@starttoc{toc}} \parskip0pt \parindent1em \def\Panel#1#2#3#4{\multicolumn{#3}{){\columncolor{#2}}#4}{#1}} \newenvironment{reflist}{% \begin{raggedright}\begin{list}{} {% \setlength{\topsep}{0pt}% \setlength{\rightmargin}{0.25in}% \setlength{\itemsep}{0pt}% \setlength{\itemindent}{0pt}% \setlength{\parskip}{0pt}% \setlength{\parsep}{2pt}% \def\makelabel##1{\itshape ##1}}% } {\end{list}\end{raggedright}} \newenvironment{sansreflist}{% \begin{raggedright}\begin{list}{} {% \setlength{\topsep}{0pt}% \setlength{\rightmargin}{0.25in}% \setlength{\itemindent}{0pt}% \setlength{\parskip}{0pt}% \setlength{\itemsep}{0pt}% \setlength{\parsep}{2pt}% \def\makelabel##1{\upshape ##1}}% } {\end{list}\end{raggedright}} \newenvironment{specHead}[2]% {\vspace{20pt}\hrule\vspace{10pt}% \phantomsection\label{#1}\markright{#2}% \pdfbookmark[2]{#2}{#1}% \hspace{-0.75in}{\bfseries\fontsize{16pt}{18pt}\selectfont#2}% }{} \def\TheFullDate{2012-01-15 (revised: 15 January 2012)} \def\TheID{\makeatother } \def\TheDate{2012-01-15} \title{Encapsulation of Soft Computing Approaches within Itemset Mining -A Survey} \author{}\makeatletter \makeatletter \newcommand*{\cleartoleftpage}{% \clearpage \if@twoside \ifodd\c@page \hbox{}\newpage \if@twocolumn \hbox{}\newpage \fi \fi \fi } \makeatother \makeatletter \thispagestyle{empty} \markright{\@title}\markboth{\@title}{\@author} \renewcommand\small{\@setfontsize\small{9pt}{11pt}\abovedisplayskip 8.5\p@ plus3\p@ minus4\p@ \belowdisplayskip \abovedisplayskip \abovedisplayshortskip \z@ plus2\p@ \belowdisplayshortskip 4\p@ plus2\p@ minus2\p@ \def\@listi{\leftmargin\leftmargini \topsep 2\p@ plus1\p@ minus1\p@ \parsep 2\p@ plus\p@ minus\p@ \itemsep 1pt} } \makeatother \fvset{frame=single,numberblanklines=false,xleftmargin=5mm,xrightmargin=5mm} \fancyhf{} \setlength{\headheight}{14pt} \fancyhead[LE]{\bfseries\leftmark} \fancyhead[RO]{\bfseries\rightmark} \fancyfoot[RO]{} \fancyfoot[CO]{\thepage} \fancyfoot[LO]{\TheID} \fancyfoot[LE]{} \fancyfoot[CE]{\thepage} \fancyfoot[RE]{\TheID} \hypersetup{citebordercolor=0.75 0.75 0.75,linkbordercolor=0.75 0.75 0.75,urlbordercolor=0.75 0.75 0.75,bookmarksnumbered=true} \fancypagestyle{plain}{\fancyhead{}\renewcommand{\headrulewidth}{0pt}} \date{} \usepackage{authblk} \providecommand{\keywords}[1] { \footnotesize \textbf{\textit{Index terms---}} #1 } \usepackage{graphicx,xcolor} \definecolor{GJBlue}{HTML}{273B81} \definecolor{GJLightBlue}{HTML}{0A9DD9} \definecolor{GJMediumGrey}{HTML}{6D6E70} \definecolor{GJLightGrey}{HTML}{929497} \renewenvironment{abstract}{% \setlength{\parindent}{0pt}\raggedright \textcolor{GJMediumGrey}{\rule{\textwidth}{2pt}} \vskip16pt \textcolor{GJBlue}{\large\bfseries\abstractname\space} }{% \vskip8pt \textcolor{GJMediumGrey}{\rule{\textwidth}{2pt}} \vskip16pt } \usepackage[absolute,overlay]{textpos} \makeatother \usepackage{lineno} \linenumbers \begin{document} \author[1]{Dr. Jyothi Pillai} \author[2]{ O.P.Vyas} \affil[1]{ Bhilai Institute of Technology, Durg, Chhattisgarh, India.} \renewcommand\Authands{ and } \date{\small \em Received: 8 December 2011 Accepted: 31 December 2011 Published: 15 January 2012} \maketitle \begin{abstract} Data Mining discovers patterns and trends by extracting knowledge from large databases.Soft Computing techniques such as fuzzy logic, neural networks, genetic algorithms, rough sets, etc. aims to reveal the tolerance for imprecision and uncertainty for achieving tractability, robustness and low-cost solutions. Fuzzy Logic and Rough sets are suitable for handling different types of uncertainty. Neural networks provide good learning and generalization. Genetic algorithms provide efficient search algorithms for selecting a model, from mixed media data. Data mining refers to information extraction while soft computing is used for information processing. For effective knowledge discovery from large databases, both Soft Computing and Data Mining can be merged. Association rule mining (ARM) and Itemset mining focus on finding most frequent item sets and corresponding association rules, extracting rare itemsets including temporal and fuzzy concepts in discovered patterns. This survey paper explores the usage of soft computing approaches in itemset utility mining. \end{abstract} \keywords{data mining, soft computing, itemset mining, fuzzy logic, neural networks, genetic algorithm.} \begin{textblock*}{18cm}(1cm,1cm) % {block width} (coords) \textcolor{GJBlue}{\LARGE Global Journals \LaTeX\ JournalKaleidoscope\texttrademark} \end{textblock*} \begin{textblock*}{18cm}(1.4cm,1.5cm) % {block width} (coords) \textcolor{GJBlue}{\footnotesize \\ Artificial Intelligence formulated this projection for compatibility purposes from the original article published at Global Journals. However, this technology is currently in beta. \emph{Therefore, kindly ignore odd layouts, missed formulae, text, tables, or figures.}} \end{textblock*} \let\tabcellsep& \section[{Introduction}]{Introduction} \section[{This paper presents a brief overview of exploration of soft computing approaches in itemset}]{This paper presents a brief overview of exploration of soft computing approaches in itemset}\par Author ? : Associate Professor, Bhilai Institute of Technology, Durg, Chhattisgarh, India. E-mail : jyothi\textunderscore rpillai@rediffmail.com Author ? : Professor, Indian Institute of Information Technology Allahabad, U.P., India. E-mail : dropvyas@gmail.com utility mining. Section 2 and Section 3 discuss theoretical definitions related to Data Mining, Itemset Utility Mining and Temporal Mining. Section 4 discusses the state of art of soft computing tools. Section 5 presents usage of different soft computing methods in data mining, itemset mining and temporal mining. Section 6 presents conclusion and future work. \section[{II.}]{II.} \section[{Data mining}]{Data mining}\par Data mining is the technique of automatic finding of hidden patterns and information elicitation from huge volume of raw data stored in data bases, data warehouses and other data repositories for making better business decisions, finding sales trends, in developing smarter marketing campaigns, and to predict customer loyalty.\par Two categories of Data mining tasks are; Descriptive Mining and Predictive Mining. The Descriptive Mining techniques include Clustering, Association Rule Discovery, and Sequential Pattern Discovery, which is used to find human-interpretable patterns that describe the data in the form of clusters, itemsets, association rules and sequential patterns. The Predictive Mining techniques such as Classification, Regression, Deviation Detection, are used to classify objects or to predict future values of other variables.\par One of the most important research areas in the field of Data mining is ARM. Association rules are used to identify relationships among a set of items in a transactional dataset. Apriori algorithm, given by Agrawal, Imielinski and Swami in 1993, is the first association rule mining algorithm, which influenced not only the association rule mining community, but also has impact on other data mining fields. Apriori and all its variants like Partition, Pincer-Search, Incremental, Border algorithm etc. take too much computer time to compute all the frequent item sets and usually consider only the frequency of items in itemsets. \section[{III.}]{III.} \section[{Itemset mining a) Frequent Itemset Mining}]{Itemset mining a) Frequent Itemset Mining}\par Frequent itemsets are itemsets that occur frequently in a transaction data set. The goal of Frequent Itemset Mining is to identify all the frequent itemsets in a transaction dataset. A frequent itemset is the itemset having frequency support greater a minimum user specified threshold {\ref [JV2011]}. ]. An emerging topic in the field of data mining is Utility Mining which is an extension of Frequent Itemset mining. The main objective of Utility Mining is to identify the itemsets with highest utilities, by considering profit, quantity, cost or other user preferences. In many real-life applications, high-utility itemsets consist of rare items also {\ref [JV2010]}. Soft computing aims to uncover the tolerance for vagueness, partial truth and approximation to achieve tractability, robustness and solutions with low cost. Soft computing methodologies consisting of fuzzy sets, neural networks, genetic algorithms, and rough sets are combined with data mining for knowledge discovery in large databases {\ref [SSP 2002}]. The resultant technique is a more intelligent system which provides a humaninterpretable, low cost solution.\par A Association Rule Mining (ARM)-The problem of mining association rules was first introduced in [RTA1993]. ARM is a popular technique for finding cooccurrences, correlations, frequent-patterns, associations among items in a set of transactions or a database. Rules with confidence and support above user-defined thresholds (minconf and minsup) were found. ARM process can be divided into two steps. The first step involves finding all frequent itemsets in databases. Next, association rules are generated from these frequent itemsets. \section[{b) Rare Itemset Mining}]{b) Rare Itemset Mining}\par The basic Bottleneck of ARM is Rare Item Problem. In many applications, some items appear very frequently in the data, while others rarely appear. In many practical situations such as security, business strategies, pattern extraction from web page access logs, biology, medicine and super market shelf management, the rare combinations of items in the itemset with high utilities provide very useful insights to the user [JV2010]. \section[{c) Utility Mining}]{c) Utility Mining}\par Identification of the itemsets with high utilities is called as Utility Mining. The frequency of itemset is not sufficient to reflect the actual utility of an itemset [JV2011]. For example, the sales executives are not interested in frequent itemsets which do not yield significant profit. Mining of high utility itemsets is one of the most challenging recent data mining tasks. The utility value of an item depends on its evaluation e.g. if cola has support 30 and profit of 3\%, cake may have support 10 but with a profit of 30\%. This indicates that the utility of cake is higher than cola.\par Utility mining model was proposed in [YHG2006] to define the utility of itemset. Utility is measured by analyzing how useful or profitable an itemset X is to user. The utility of an itemset X, u(X) is the sum of the utilities of itemset X in all the transactions containing X. An itemset X is called a high utility itemset if and only if u(X) >= min\textunderscore utility, where min\textunderscore utility is a user-defined minimum utility threshold [YHG2006]. For example, a computer system may be more profitable than a telephone in terms of profit. The main objective of high-utility itemset mining is to find all those itemsets having utility greater or equal to user-defined minimum utility threshold. \section[{IV.}]{IV.} \section[{Soft computing a) Importance of Soft Computing}]{Soft computing a) Importance of Soft Computing}\par Soft computing is tolerant of vagueness, imprecision, uncertainty, incomplete truth and approximation. The main components of Soft Computing are: fuzzy logic (FL), neural networks (NN), probabilistic reasoning (PR), genetic algorithms (GA), and chaos theory (ChT), which are summarized:-i. Fuzzy Logic [RV 2011] Lotfi Zadeh conceived the concept of FL. FL is used to deal with uncertain or vague data, considered as fuzzy sets. In FL procedure, attribute values are transformed to fuzzy values and corresponding fuzzy membership or truth values are calculated.\par ii. Neural Networks [RV 2011] NN is a network of artificial neurons which are simple processing elements which process information with a connectionist approach to computation [RAA2001]. An important property of these networks is their inductive nature, which uses "learning by example" in problems solving.\par iii. Genetic Algorithms [RV 2011] GA is a flexible, heuristic and inductive search technique based on the theory of natural selection. GA learning consists of following steps: An initial input is created which consists of randomly generated rules. Each rule is represented using a string of bits. The fitness of a rule is evaluated by its classification accuracy on the training samples set. This process of generating new populations based on previous populations of rules is repeated till each rule of a population satisfies a pre defined fitness threshold [RAA2001]. vi. Chaos Theory [RAA2001]. A chaotic system is a deterministic system that exhibits random behavior. ChT deals with the non-linear dynamical systems that exhibit extreme sensitivity to initial conditions. \section[{b) Need of Soft Computing in Data Mining}]{b) Need of Soft Computing in Data Mining}\par By incorporation of Soft Computing, there is a significant increase in effectiveness of artificial intelligence systems. All techniques have their own uniqueness based upon which they can be properly used in data mining process. iii. Genetic Algorithm in Data mining GA processing objects operate directly to set, queue, matrices, charts, and other structure. GA adopts probability rules to lead search direction. Genetic programming concepts have been used for developing Knowledge discovery systems. For better attribute interaction, GAs can be used. iv. \section[{Rough Sets in Data mining}]{Rough Sets in Data mining}\par The main aim of RS is stimulation of approximation of concepts. Mathematical tools are offered by RS to extract hidden patterns in data and therefore are used in data mining. In data mining, RS can be used as a framework where precise data is not necessary and in the areas where approximate data is of great help. In data processing RST can be used for computing lower and upper approximation [RV 2011]. v. \section[{Probabilistic Reasoning in Data mining}]{Probabilistic Reasoning in Data mining}\par Statistics or Probabilistic Theory forms a basis for good management and also plays a very important role in the data mining methods [RAA2001].\par vi. \section[{Chaos Theory in Data mining}]{Chaos Theory in Data mining}\par The predictability can be done using chaotic analysis and also prediction strategies of system's behavior can be formulated. ChT deals efficiently with noisy nonlinear systems. Chaotic computing gives a tool to determine a new perspective of nonlinear data analysis [RAA2001].\par V. \section[{Literature survey a) Application of Soft Computing in Data Mining}]{Literature survey a) Application of Soft Computing in Data Mining}\par By combining the advantages of both Data mining and soft computing paradigms, the techniques can be used for discovering knowledge in databases. In this section, a literature survey of integration of various soft computing methodologies and data mining is presented.\par i. \section[{Fuzzy Logic}]{Fuzzy Logic}\par In retrieval of information, the main complexity is identifying relevant information, i.e. the nearest or the most similar according to user's need or expectation. This problem motivated to use fuzzy sets in knowledge representation thus enabling the user to express his prospect in a language not far from natural. Another reason is the approximate matching between the user's requirements and existing values in the database, on the basis of similarities and degrees of satisfiability.\par The thesis report of Jianxiong Luo [JL1999] explores integrating FL with two data mining methods (association rules and frequency episodes) for intrusion detection. In intrusion detection, many quantitative features are involved and also security is fuzzy.\par Au and Chan [WK1999] use an adjusted difference between experimental and probable frequency counts of attributes for finding out fuzzy association rules in relational datasets. The algorithm discovers both positive and negative rules and is able to cope with fuzzy class boundaries and missing values.\par The authors in [BDLMR2007] focus on the applications of fuzzy techniques for information retrieval and data mining in real-world situations such as medical, educational, chemical and multimedia have been illustrated.\par In real-time systems, for example in e-banking, assessing and determining any phishing websites is a complex and dynamic problem because of ambiguities involved. Aburrous et al present a intelligent, flexible and efficient system approach to deal with 'fuzziness' in the e-banking phishing website using fuzzy data mining techniques [AHDT2010].\par In [KMA2012], the authors present an overview of the applications of fuzzy decision tree in heterogeneous fields. It is used dynamically in various fields such as intrusion detection, querying processes, cognitive process analysis (Human Computer Interaction), biometrics authentication, stock-market, parallel processing support, information retrieval and also in data mining.\par ii. \section[{Neural Network}]{Neural Network}\par The paper [HRH1996] presents a method to find out symbolic classification rules using NNs. By the application of FL, the system's performance improved for diagnosing diabetes in patients.\par Roohollah Etemadi et al propose a GA approach based on k-means clustering algorithm which can select cluster centers in a better manner [R2012]. All data objects are firstly clustered through k-means algorithm. Secondly, for each data object a pattern is generated by considering the generated clusters. On comparing with other related algorithms, the authors state that the proposed algorithm is more efficient than k-means algorithm and other algorithms. According to authors, the proposed algorithm achieves better and accurate predictive results as compared to other two competent learners. iv. \section[{Rough Set}]{Rough Set}\par RST deals with classificatory study of information systems. Z. Pawlak proposed this mathematical approach which is a powerful tool for dealing with vague data. Using RS method without deteriorating the quality of approximation, minimal attribute sets, and minimal length decision rules corresponding to lower or upper approximation can be extracted [W2012].\par Prasanta et al proposed an approach based on RST which mine concise rules from inconsistent data [PRBB2011]. Firstly, lower and upper approximation is computed for each concept. Then a learning algorithm is adopted for building classification rules for each concept which satisfies classification accuracy. Test results show that the approach produced effective and minimal rules and offers more accurate results applied on several real life datasets.\par In many fields such as inductive reasoning, classification, pattern recognition, cluster analysis, automatic learning algorithms, RST plays a significant role and is used in different domains like Medicine, Banking, Marketing and Engineering. In [S2011], A.S. Salama described some topological properties of RS which will help get rich results and discover hidden relations between data and also help in producing accurate programs.\par Abdul Nassar proposes that using RST concept, clusters can be generated without any additional information for example probability distribution or fuzzy membership function {\ref [A2011]}. By considering Lower approximation important rules of the target set can be generated. A reduct rule set of high importance can be generated by considering generated rules as attributes and a new decision table can be constructed.\par Wen-Yau proposed a clustering technique which uses GA and RST [W2012]. After clustering, Apriori algorithm is used to discover association rules between products of same cluster and then marketing people can suggest related products to the targeting group. RS is used to generate rules and these rules are applied to various GA parts. \section[{b) Application of Soft Computing in Itemset Mining}]{b) Application of Soft Computing in Itemset Mining}\par A literature survey of exploration of different soft computing approaches in itemset mining is discussed in this section. i. \section[{Fuzzy Logic}]{Fuzzy Logic}\par Wai-Ho introduced a novel technique, called FARM (Fuzzy Association Rule Miner) to mine fuzzy association rules [WK1999] which uses linguistic terms for representing revealed regularities and exceptions, based on fuzzy set theory. The rules generated are called fuzzy association rules. FARM also discovers interesting associations between different quantitative values. One more advantage of FARM is that it can reveal both positive and negative association rules. A positive association rule indicates presence of another attribute value along with a certain attribute value whereas a negative association rule indicates absence of another attribute value along with a certain attribute value. Wai-Ho et al discuss that experimental results show FARM to be capable of discovering meaningful and useful fuzzy association rules.\par Yi-Chung Hu et al proposed a learning algorithm, which acts as a knowledge acquisition tool for classification problems to efficiently generate fuzzy association rules [YRG2002]. In first phase, from training samples, large fuzzy grids are generated by fuzzy partitioning of each attribute and in second phase, for classification problems, fuzzy association rules by large fuzzy grids are generated. Experimental results on iris data indicate that the proposed algorithm accurately derive fuzzy association rules for classification problems.\par One of the most essential areas of the application of fuzzy set theory is Fuzzy rule-based systems [CMM2004]. The advantages of using fuzzy systems for knowledge discovery processes are; information dealing with uncertain data, considering multi-variable relationships; human understandable results, easy information modification by an expert, easy adaptability to the given problem and high automated process. Fuzzy systems improve the interpretation and understandability of consumer models. In Using fuzzy C-Means algorithm, Quantitative attributes are divided into several fuzzy sets and accordingly membership values are generated. Then a supervised association rule algorithm is employed for discovering interesting FARs. Generated Fuzzy rules are used to build classification system. C4.5, Naïve Bayes, and ID3 classifiers are used for classification and accordingly fuzzy classified association rules are discovered. The authors discuss that the number of generated rules is reduced due to the usage of fuzzy linguistic values.\par et al presented a novel approach to mine weighted FARs effectively and address the issue of invalidation of downward closure property (DCP) in weighted ARM, where each item is assigned a weight according to its significance with respect to some user defined criteria [MSC2010]. Prakash et al present a qualitative fuzzy ARM (FARM) approach for mining FARs for the quantitative attributes [PP2011]. The authors evaluated the performance of qualitative FARM by experimenting with real data sets. Results prove that the qualitative approach discover more accurate association rules in less time with increased execution speed.\par A novel approach is presented by Vedula Venkateswara Rao et al in [VES2012] for effectively mining frequent Item sets and generating association rules (ARs) based on fuzzy Apriori and weighted fuzzy Apriori. In weighted association rule mining (WARM), each item is assigned a weight with respect to its importance to some user defined criteria. Both binary data and fuzzy data are used in the proposed approach and Frequent Item Sets are generated. The Fuzzy Apriori algorithm (Apriori-Total) proposed in [VES2012] is founded on a tree structure called the T-tree to store frequent item set information.\par K. Suriya Prabha et al proposed an approach that integrates FL and tree-based algorithm. The approach constructs a compact sub-tree for finding fuzzy frequent item [SL2012]. The authors conclude that the presented approach is quite efficient than other algorithms when evaluated on the basis of execution time, memory usages and search space for generating fuzzy frequent itemsets.\par Ferdinando et al present a novel method for detecting association rules from datasets based on fuzzy transforms [FS2012]. AprioriGen algorithm is used for extracting fuzzy association rules which are represented in the form of linguistic expressions. A preprocessing phase is performed for determining optimal fuzzy partition of quantitative attributes domains.\par Roohollah Etemadi states that one of the most well-known clustering methods is K-means algorithm which forms the base for other clustering approaches [R2012]. K-means and k-methods are heuristic partitioning algorithms where as Fuzzy k-means and Fuzzy k-methods are equivalent fuzzy type algorithms. In these partitioning methods, firstly k number of partitions is generated from data where each partition will contain at least one data. If crisp partitioning is performed, then a particular data will be present in only one cluster but if fuzzy partitioning is assumed then a particular data may be present in different clusters.\par ii. Anjana Pandey et al proposed an algorithm RS Model for Discovering Hybrid Association Rules [RSHAR] algorithm, for mining hybrid association rules using rough set approach [AP2009]. In RSHAR algorithm, the participant tables are combined into a general table for generating rules to express the relationship between two or more domains belonging to different database tables and then on selected dimension, the mapping code is applied. Then frequent itemsets are generated through equivalence classes and also the mapping code is transformed into real dimensions. \section[{Neural}]{Neural}\par In [DHM2002], Daniel Delic et al emphasis on the comparison of association rules procedure and rough sets procedure. The proposed association rules method focus on the analysis of data bases containing boolean-valued attributes only. The authors conclude that there is a considerable reduction in computing time in the rough set algorithm.\par A. Anitha et al proposed to combine upper approximation based rough set clustering and Apriori selective ARM for e-learning recommendation [AK2011].\par In making e-learning recommendations, similar learning patterns are considered instead of all clicks stream sequences. The proposed algorithm resulted in dense clusters with less computational complexity and reduced number of extracted rules, which are highly relevant and meaningful. \section[{VI.}]{VI.} \section[{Conclusion}]{Conclusion}\par There has been substantial commercial interest as well as active research in data mining area for developing new and improved approaches for extracting information, relationships, and patterns from large datasets. Soft computing may be viewed as a foundation component for the emerging field of conceptual intelligence [RAA2001]. Hence Soft computing techniques can be encapsulated in Data mining for knowledge discovery in large databases. This paper presents a brief overview of various soft computing approaches used in itemset mining.\par In future we will incorporate soft computing methodologies and itemset mining for mining high utility itemsets. \begin{figure}[htbp] \noindent\textbf{}\includegraphics[]{image-2.png} \caption{\label{fig_0}}\end{figure} \begin{figure}[htbp] \noindent\textbf{} \par \begin{longtable}{P{0.39279674796747965\textwidth}P{0.0019349593495934958\textwidth}P{0.0019349593495934958\textwidth}P{0.0027642276422764228\textwidth}P{0.0022113821138211383\textwidth}P{0.43343089430894305\textwidth}P{0.0038699186991869917\textwidth}P{0.0022113821138211383\textwidth}P{0.0011056910569105691\textwidth}P{0.0013821138211382114\textwidth}P{0.006357723577235772\textwidth}} \multicolumn{5}{l}{Hongjun Lu et al propose an approach which can}\tabcellsep \multicolumn{5}{l}{to learn new fraud patterns as the types of fraud}\\ \multicolumn{5}{l}{extract concise symbolic rules accurately using NN. The}\tabcellsep evolved.\tabcellsep \tabcellsep \tabcellsep \tabcellsep \\ \multicolumn{5}{l}{NN is trained for achieving required accuracy rate. Then}\tabcellsep \multicolumn{5}{l}{Madhusmita Swain et al introduced NNs for}\\ through\tabcellsep network\tabcellsep pruning\tabcellsep algorithm,\tabcellsep repeated\tabcellsep simplifying\tabcellsep classification\tabcellsep problem,\tabcellsep IRIS\tabcellsep plant\\ \multicolumn{5}{l}{connections of the Anuj et al discuss that it is more expensive to connect a new customer than to maintain an existing loyal customer [AP2011]. The authors propose a NN based approach for predicting customer churn in cellular wireless services subscription and conclude a promising solution for customer churn management. The experimental results show that NN based method can predict customer churn with more than 92\%}\tabcellsep \multicolumn{5}{l}{classification [MSSA2012]. The problem identifies IRIS plant species on basis of plant attribute measurements. The authors used back propagation learning algorithm to train Multilayer feed-forward networks for identification of IRIS plants based on measurements such as length and width of sepal and length and width of petal. The authors conclude that Multi Layer Feed Forward NN (MLFF) is faster in terms of learning and is more accurate. iii. Genetic Algorithm family of computational models which are inspired by evolution are GAs [E2011]. GA implementation begins with a population of random chromosomes. To create next generation of chromosomes from current population, GA uses three main types of rules: 1. The individuals called parents are selected through Selection rules, which contribute to next generation population. 2. Two parents are combined using Crossover rules to form next generation children. 3. Random changes are applied to individual parents using Mutation rules for forming children. Ramesh Kumar et al presented a novel algorithm for rule prioritizing, which are generated by apriori algorithm through GA [RI2011]. E.P. Ephzibah proposed a new way to improve the performance of a model by combining GAs and FL, for feature selection and classification [E2011]. The proposed automated pattern classification system identifies and selects a subset of pattern from a larger set of features using fuzzy rule-based classification system.}\tabcellsep ( D D D D ) C 2012 Year\\ accuracy.\tabcellsep \tabcellsep \tabcellsep \tabcellsep \tabcellsep \tabcellsep \tabcellsep \tabcellsep \tabcellsep \\ \multicolumn{5}{l}{Kamruzzaman et al propose a novel four-phase}\tabcellsep \tabcellsep \tabcellsep \tabcellsep \tabcellsep \\ \multicolumn{5}{l}{data mining algorithm using ANNs, referred as ESRNN}\tabcellsep \tabcellsep \tabcellsep \tabcellsep \tabcellsep \\ \multicolumn{5}{l}{(Extraction of Symbolic Rules from ANNs), for extracting}\tabcellsep \tabcellsep \tabcellsep \tabcellsep \tabcellsep \\ \multicolumn{5}{l}{symbolic rules [KJ2011]. The algorithm uses back}\tabcellsep \tabcellsep \tabcellsep \tabcellsep \tabcellsep \\ \multicolumn{5}{l}{propagation learning. Network architecture is defined}\tabcellsep \tabcellsep \tabcellsep \tabcellsep \tabcellsep \\ \multicolumn{5}{l}{and refined in the first phase and second phases. By}\tabcellsep \tabcellsep \tabcellsep \tabcellsep \tabcellsep \\ \multicolumn{5}{l}{using heuristic clustering algorithm, the nodes in hidden}\tabcellsep \tabcellsep \tabcellsep \tabcellsep \tabcellsep \\ \multicolumn{5}{l}{layers are discretized in third phase. Then symbolic}\tabcellsep \tabcellsep \tabcellsep \tabcellsep \tabcellsep \\ \multicolumn{5}{l}{rules are extracted from frequent patterns using}\tabcellsep \tabcellsep \tabcellsep \tabcellsep \tabcellsep \\ \multicolumn{2}{l}{extraction algorithm.}\tabcellsep \tabcellsep \tabcellsep \tabcellsep \tabcellsep \tabcellsep \tabcellsep \tabcellsep \\ \multicolumn{5}{l}{Mohammad Iquebal Akhter et al discusses in}\tabcellsep \tabcellsep \tabcellsep \tabcellsep \tabcellsep \\ \multicolumn{5}{l}{detail the function of ANN in preventing fraud in}\tabcellsep \tabcellsep \tabcellsep \tabcellsep \tabcellsep \\ \multicolumn{5}{l}{telecommunication services [MM2012]. A Fraud}\tabcellsep \tabcellsep \tabcellsep \tabcellsep \tabcellsep \\ \multicolumn{5}{l}{Detection System using ANN gathers historical data}\tabcellsep \tabcellsep \tabcellsep \tabcellsep \tabcellsep \\ \multicolumn{5}{l}{which is preprocessed and is used for training the NN}\tabcellsep \tabcellsep \tabcellsep \tabcellsep \tabcellsep \\ \multicolumn{5}{l}{for building a model which incorporates frequent fraud}\tabcellsep \tabcellsep \tabcellsep \tabcellsep \tabcellsep \\ \multicolumn{5}{l}{patterns. Finally, the model is applied to new business}\tabcellsep \tabcellsep \tabcellsep \tabcellsep \tabcellsep \end{longtable} \par \caption{\label{tab_0}}\end{figure} \begin{figure}[htbp] \noindent\textbf{} \par \begin{longtable}{P{0.85\textwidth}} Networks\\ {}[VI2012] NNs have the capability to interpret\\ meaning from complicated or vague data and hence\\ can be used for extracting patterns and detecting trends\\ which are difficult for humans or other computer\\ techniques to notice\\ P. Sermswatsri et al proposes a more efficient\\ method of frequent pattern mining by using Associative\\ Classification method and NN [PS2006]. The proposed\\ NN Associative Classification (NAC) method can be\\ used to build more accurate and efficient classifiers. The\\ authors conclude that the experimental results of NAC\\ on datasets show improved accuracy rates.\\ Divya Bhatnagar et al propose an efficient\\ technique for frequent itemsets mining in large\\ databases using Optical NN Model [DNS2011]. The\\ proposed technique removes the need for generating\\ candidate sets for ARM to find frequent itemsets. The\\ time complexity and space complexity of this technique\\ is very low as optical NN can perform several\\ computations simultaneously.\\ In [ASP2011], an efficient algorithm named Multi\\ Level Feed Forward Mining (MLFM) is proposed by Amit\\ Bhagat et al, for mining of multiple-level association\\ rules efficiently from large transaction databases.\end{longtable} \par \caption{\label{tab_1}}\end{figure} \footnote{© 2012 Global Journals Inc. (US)Global Journal of Computer Science and Technology} \footnote{© 2012 Global Journals Inc. (US)} \backmatter \subsection[{Year}]{Year}\par Maybin Muyeba by scanning the database only once, and thus produces fast output NN Associative Classification system proposed by Prachitee B. Shekhawat builds a classifier with the help of Back propagation NN [PS2011].\par iii. \subsection[{Genetic Algorithm}]{Genetic Algorithm}\par Xiaowei Yan et al designed an evolutionary mining strategy based on a GA called ARMGA model \hyperref[b20]{[XCS2007]}. The authors discuss that, ARMGA model is efficient for global searching when search space is very large. Generally for rules mining, GAs are classified into two categories, according to encoding of rules in the population of chromosomes. In one encoding method called the Michigan Approach, each rule is encoded into an individual. In another method referred as the Pittsburgh Approach, the set of rules are encoded into a chromosome. ARMGA model is based on the Michigan strategy, where each association rule is encoded in a single chromosome.\par In [ACC2007], Ansaf Salleb-Aouissi et al proposed QUANTMINER, a mining quantitative association rules system, which is based on GA that dynamically discovers "good" intervals in association rules.\par Peter \begin{bibitemlist}{1} \bibitem[Anitha and Krishnan (2011)]{b1}\label{b1} ‘A Dynamic Web Mining Framework for E-Learning Recommendations using Rough Sets and Association Rule Mining’. A Anitha , N Krishnan . \textit{International Journal of Computer Applications} January 2011. 12 (11) p. . \bibitem[Basheer et al. (2012)]{b8}\label{b8} ‘A Genetic Algorithm for Discovering Classification Rules in Data Mining’. M Basheer , Hamid Al-Maqaleh , Shahbazkia . \textit{International Journal of Computer Applications} March 2012. 41 (18) p. . \bibitem[Liang ()]{b18}\label{b18} \textit{A Hybrid Evolutionary Computing For Market Segmentation, Business and Information}, Wen-Yau Liang . 2012. July 3-5. Sapporo. p. . \bibitem[Sharma and Kumar Panigrahi (2011)]{b3}\label{b3} ‘A Neural Network based Approach for Predicting Customer Churn in Cellular Network Services’. Anuj Sharma , Prabin Kumar Panigrahi . \textit{International Journal of Computer Applications} August 2011. 27 (11) p. . \bibitem[Ma and Tang ()]{b15}\label{b15} ‘A New Reduction Implementation Based on Concept’. Tinghuai Ma , Meili Tang . \textit{Journal of Information and Computing Science} 1746-7659. 2006. 1 (5) p. . \bibitem[Abdul Nassar et al. ()]{b0}\label{b0} ‘Application of Rough Sets in Data Mining, A Project Report Submitted in partial fulfillment of the requirements for the award of the degree of Master of Technology in Computer Science and Engineering’. . A Abdul Nassar , M A A ; M. Aburrous , K Hossain , F Dahal , Thabtah . \textit{Intelligent phishing detection system for e-banking using fuzzy data mining}, 2011. 2010. Elsevier. 37 p. . \bibitem[Xcs2007] Xiaowei Yan et al. ()]{b20}\label{b20} ‘ARMGA: Identifying Interesting Association Rules With Genetic Algorithms’. [ Xcs2007] Xiaowei Yan , Chengqi Zhang , Shichao Zhang . \textit{Applied Artificial Intelligence} 0883-9514 print/1087-6545 online. 2007. 19 p. . \bibitem[Mitra et al. ()]{b13}\label{b13} ‘Data Mining in Soft Computing Framework: A Survey’. Sushmita Mitra , K Sankar , Pabitra Pal , Mitra . \textit{IEEE Transactions On Neural Networks} JANUARY 2002. 13 (1) p. . \bibitem[Ghosh and Biswas (2010)]{b12}\label{b12} ‘Debasree Sarkar, Partha Pratim Sarkar, Mining Frequent Itemsets Using Genetic Algorithm’. Soumadip Ghosh , Sushanta Biswas . \textit{International Journal of Artificial Intelligence \& Applications (IJAIA)} October 2010. 1 (4) p. . \bibitem[Mangalampalli and Pudi ()]{b6}\label{b6} ‘Fuzzy Association Rule Mining Algorithm for Fast and Efficient Performance on Very Large Datasets, Fuzzy Systems, FUZZ-IEEE’. Ashish Mangalampalli , Vikram Pudi . \textit{IEEE International Conference on Fuzzy Systems (FUZZ-IEEE)}, (Jeju Island, Korea) 2009. 20-24 Aug., 2009. p. . \bibitem[[cmm2004 et al. (2004)]{b9}\label{b9} ‘Fuzzy Association Rules For Estimating Consumer Behaviour Models And Their Application To Explaining Trust In Internet Shopping’. ] J [cmm2004 , F J Casillas , F J Martínez-López , Martínez . \textit{Fuzzy Economic Review} Nov. 2004. (9) p. . \bibitem[Xcs2009] Xiaowei Yan et al. ()]{b21}\label{b21} ‘Genetic algorithm-based strategy for identifying association rules without specifying actual minimum support’. [ Xcs2009] Xiaowei Yan , Chengqi Zhang , Shichao Zhang . \url{www.sciencedirect.com} \textit{Expert Systems with Applications} 2009. 36 p. . \bibitem[Vi2012] Vinodhini and Katiyaar ()]{b17}\label{b17} ‘Ina Kapoor Sharma Use of Data Mining \& Neural Network in’. [ Vi2012] Vinodhini , Katiyaar . \textit{Commercial Application International Journal of Computer Science and Information Technologies (IJCSIT)} 0975-9646. 2012. 3 (3) p. . \bibitem[Delic et al. ()]{b10}\label{b10} ‘Mattis Neiling, Rough Sets and Association Rules -which is efficient?’. Daniel Delic , -J Hans , Lenz . \textit{14th Conference on Computational Statistics (CompStat2002)}, (Berlin, HU) August 24-28, 2002. p. . \bibitem[Prakash et al. ()]{b16}\label{b16} ‘Mining Frequent Itemsets from Large Data Sets using Genetic Algorithms, IJCA Special Issue on’. R Prakash , S S V N Govardhan , Sharma . \textit{Artificial Intelligence Techniques -Novel Approaches \& Practical Applications}, 2011. AIT. p. . \bibitem[Dns2011] Divya et al. ()]{b11}\label{b11} ‘Mining Frequent Itemsets without Candidate Generation using Optical Neural Network, IJCA Special Issue on’. [ Dns2011] Divya , Neeru Bhatnagar , A S Adlakha , Maybin Saxena ; M. Sulaiman Khan , Muyeba . \textit{Artificial} 0302-9743. 2006. 46 p. . (Christos Tjortjis, Frans Coenen, An effective Fuzzy Healthy Association Rule Mining Algorithm (FHARM)) \bibitem[Bhagat et al. (2011)]{b5}\label{b5} ‘Ontological Frequent Patterns Mining by potential use of Neural Network’. Amit Bhagat , Sanjay Sharma , K R Pardasani . \textit{International Journal of Computer Applications} December 2011. 36 (10) p. . \bibitem[Anandhavalli et al. ()]{b4}\label{b4} ‘Optimized association rule mining using genetic algorithm’. M Anandhavalli , Ayush Suraj Kumar Sudhanshu , Kumar , M K Ghose . \textit{Advances in Information Mining}, 2009. Bioinfo Publications. 1 p. . \bibitem[Bdlmr2007] B. Bouchon-Meunier et al. ()]{b7}\label{b7} \textit{Real world fuzzy logic applications in data mining and information retrieval}, [ Bdlmr2007] B. Bouchon-Meunier , M Detyniecki , M.-J Lesot , C Marsala , M Rifqi . 2007. Springer. p. . \bibitem[Ni ()]{b19}\label{b19} ‘Research of Data Mining Based on Neural Networks’. Xianjun Ni . \textit{World Academy of Science, Engineering and Technology} 2008. 39. \bibitem[Pandey and Pardasani (2009)]{b2}\label{b2} \textit{Rough Set Model for Discovering Hybrid Association Rules}, Anjana Pandey , K R Pardasani . June 2009. 2. \bibitem[Lin ()]{b14}\label{b14} ‘Rough Set Theory in Very Large Databases’. T Y Lin . \textit{Proceedings of Symposium on Modeling, Analysis and Simulation, IMACS Multiconference (Computational Engineering in Systems Applications}, (Symposium on Modeling, Analysis and Simulation, IMACS Multiconference (Computational Engineering in Systems ApplicationsLille, France) July 9-12, 1996. 2 p. . \end{bibitemlist} \end{document}