Download Intelligence decision systems in enterprise information management

The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1741-0398.htm
JEIM
24,4
Intelligence decision systems in
enterprise information
management
360
Cengiz Kahraman
Received 12 September 2010
Revised 22 October 2010
8 January 2011
Accepted 8 February 2011
Department of Industrial Engineering, Istanbul Technical University,
İstanbul, Turkey
İhsan Kaya
Department of Industrial Engineering, Yıldız Technical University,
İstanbul, Turkey, and
Emre Çevikcan
Department of Industrial Engineering, Istanbul Technical University,
İstanbul, Turkey
Abstract
Purpose – The purpose of this paper is to show how intelligence techniques have been used in
information management systems.
Design/methodology/approach – The results of a literature review on intelligence decision
systems used in enterprise information management are analyzed. The intelligence techniques used in
enterprise information management are briefly summarized.
Findings – Intelligence techniques are rapidly emerging as new tools in information management
systems. Especially, intelligence techniques can be used to utilize the decision process of enterprises
information management. These techniques can increase sensitiveness, flexibility and accuracy of
information management systems. The hybrid systems that contain two or more intelligence
techniques will be more used in the future.
Originality/value – The intelligence decision systems are briefly introduced and then a literature
review is given to show how intelligence techniques have been used in information management
systems.
Keywords Intelligence techniques, Decision support systems, Enterprise information management,
Information management
Paper type Literature review
Journal of Enterprise Information
Management
Vol. 24 No. 4, 2011
pp. 360-379
q Emerald Group Publishing Limited
1741-0398
DOI 10.1108/17410391111148594
1. Introduction
Information plays a great role in the enterprise. It is consistently a hot topic for
information management, and all the time, evaluating its efficiency is a critical task.
For a corporation, there are thousands pieces of vertical and horizontal information
flows within it. The information management activities, in fact, can be regarded as the
control on those flows (Li et al., 2003). In response to global competition, enterprises are
increasingly employing information technology to conduct business electronically.
Thus, various information systems, such as enterprise resource planning (ERP),
supply chain management (SCM) and customer relationship management (CRM), are
increasingly used to gather business transaction, supplier and customer data. Those
various operational and transaction data can be transformed into information and then
knowledge by using business intelligence (BI) tools. Enterprises decision makers make
better business decisions (Wu, 2010). Decision making is one of the most important
tasks for enterprise managers, and is generally based on various data sources derived
from information systems, such as ERP, SCM, human resource management (HRM),
financial management (FM) and CRM. Numerous intelligence decision systems (IDS)
have been developed to support decision making for enterprises. A definition of
intelligence which is needed to be defined to research in a field termed artificial
intelligence and computational intelligence has only rarely been provided, and the
definitions in the literature have often been of little operational value. The intelligence
was defined as the ability of solving the hard problems. However there are basic
questions in this definition, such as how hard the problem is or who decides which
problem is hard (Chellapilla, Fogel, 1999).
Intelligent decision support systems (IDSS) is a term that describes decision support
systems that make extensive use of artificial intelligence (AI) techniques. Use of AI
techniques in management information systems has a long history, indeed terms such
as Knowledge-based systems (KBS) and intelligent systems have been used since the
early 1980s to describe components of management systems, but the term “Intelligent
decision support system“ is thought to originate the late 1970s (Holsapple and
Whinston, 1987). Flexible manufacturing systems (FMS) (Chan et al., 2000) can also be
considered examples of intelligent decision support systems. Ideally, an intelligent
decision support system should behave like a human consultant; supporting decision
makers by gathering and analysing evidence, identifying and diagnosing problems,
proposing possible courses of action and evaluating the proposed actions. The aim of
the AI techniques embedded in an intelligent decision support system is to enable these
tasks to be performed by a computer, whilst emulating human capabilities as closely as
possible. Many IDSS implementations are based on expert systems, a well established
type of KBS that encode the cognitive behaviors of human experts using predicate
logic rules and have been shown to perform better than the original human experts in
some circumstances (Baron, 1998; Turban et al., 2009). Expert systems emerged as
practical applications in the 1980s based on research in artificial intelligence performed
during the late 1960s and early 1970s. They typically combine knowledge of a
particular application domain with an inference capability to enable the system to
propose decisions or diagnoses. ( Jackson, 1998). Some research in intelligence
techniques, focused on enabling systems to respond to novelty and uncertainty in more
flexible ways is starting to be used in intelligent decision support systems. For
example intelligent agents that perform complex cognitive tasks without any need for
human intervention have been used in a range of decision support applications.
Capabilities of these intelligent agents include knowledge sharing, machine learning,
data mining, and automated inference. A range of intelligence techniques such as case
based reasoning, rough sets and fuzzy logic have also been used to enable decision
support systems to perform better in uncertain conditions (Sugumaran, 2007; Bui, and
Lee, 1999). So they can be used to construct intelligence decision systems in enterprise
information management. Enterprise information management systems in many
companies have been developed following the needs arising from administration,
control, reporting and transaction management.
Intelligence
decision systems
361
JEIM
24,4
362
In this paper, the results of a literature review on intelligence decision systems used
in enterprise information management are analyzed. The intelligence techniques used
in enterprise information management (EIM) such as fuzzy set theory (FST),
multi-agent systems (MAS), neural networks (NNs), genetic algorithms (GAs), ant
colony optimization(ACO) and particle swarm optimization (PSO) are briefly
summarized in Section 2. The results of the literature review are analyzed in Section
3. The conclusions and future research directions are discussed in Section 4.
2. Intelligence techniques
In this section, the intelligence techniques used in EIM are summarized briefly.
2.1 Multi-agent system
Multi-agents systems (MAS) theory can be viewed as an evolution of artificial
intelligence, in order to achieve autonomous computational systems. Although the
agent definition has been argued into the researcher’s community of the distributed
artificial intelligence, it is accorded that the autonomy is the main characteristic
describing an agent, the autonomy being the ability to accomplish a task and reach its
objectives without human, or any other, assistance. Particularly, each agent has got
properties as autonomy, mobility, rationality, sociability and reactivity. Furthermore,
they may have built-in reasoning mechanism permitting to propose intelligent
solutions and to evolve by experiences. Thus, intelligent agents have attributes and
methods, like the object-oriented programming, and also they have built-in beliefs,
desires and intentions, which are linked with their environment and provide them of
the states that determine their behavior The most important agent’s characteristics are:
.
autonomy;
.
communication;
.
reactivity;
.
intelligence; and
.
mobility (Cerrada et al., 2007).
2.2 Fuzzy set theory
Zadeh published the first paper, called “fuzzy sets”, on the theory of fuzzy logic in 1965.
He characterized non-probabilistic uncertainties and provides a methodology, fuzzy set
theory, for representing and computing data and information that are uncertain and
imprecise. Zadeh (1996) defined the main contribution of fuzzy logic as “a methodology
for computing with words” and pointed out two major necessities for computing with
words: “First, computing with words is a necessity when the available information is
too imprecise to justify the use of numbers, and second, when there is a tolerance for
imprecision which can be exploited to achieve tractability, robustness, low solution
cost, and better rapport with reality”. While the complexity arise and the precision is
receded, linguistic variables has been used to modeling. Linguistic variables are
variables which are defined by words or sentences instead of numbers (Zadeh, 1975).
Many problems in the real world deal with uncertain and imprecise data so
conventional approaches cannot be effective in finding the best solution. To cope with
this uncertainty, fuzzy set theory has been developed as an effective mathematical
algebra under vague environment. Although humans are comparatively efficient in
qualitative forecasting, they are unsuccessful in making quantitative predictions. Since
fuzzy linguistic models permit the translation of verbal expressions into numerical
ones, thereby dealing quantitatively with imprecision in the expression of the
importance of each criterion, some methods based on fuzzy relations are used. When
the system involves human subjectivity, fuzzy algebra provides a mathematical
framework for integrating imprecision and vagueness into the models (Kaya and Çınar,
2008). Uncertainties can be reflected in mathematical background by fuzzy sets. Fuzzy
sets have reasonable differences with crisp (classical) sets. Crisp set A in a universe U
can be defined by listing all of its elements denoted x. Unlike crisp sets, a fuzzy set à in
universe of U is defined by a membership function mÃ(x) which takes on values in the
interval [0, 1]. The definition of a fuzzy set is the extended version of a crisp set. While
the membership function can take the value of 0 or 1 in crisp sets, it takes a value in
interval [0, 1] in fuzzy sets. A fuzzy set, Ã, is completely characterized by the set of
ordered pairs ( Jahanshahloo et al., 2006):
~ ¼ x; m ~ ðxÞ jx [ X
A
ð1Þ
A
There is a huge amount of literature on fuzzy logic and fuzzy set theory. In recent
studies, the fuzzy set theory has been concerned with engineering applications. Certain
types of uncertainties are encountered in a variety of areas and the fuzzy set theory has
pointed out to be very efficient to consider these (Kaya and Çınar, 2008).
2.3 Artificial neural networks
Artificial neural networks (ANN) are optimization algorithms which are recently being
used in variety of applications with great success. Artificial neural network (ANN)
models take inspiration from the basic framework of the brain. The main parts of a
neuron are cell body, axon and dendrites. A signal transport from axon to dendrites
and passing through other neurons by synapses (Fu, 1994). ANN consists of many
nodes and connecting synapses. Nodes operate in parallel and communicate with each
other through connecting synapses. ANN is used effectively for pattern recognition
and regression. In recent years, experts prefer ANN over classical statistical methods
as a forecasting model. This increasing interest can be explained by some basic
properties of ANN. Extrapolating from historical data to generate forecasts, solving the
complex nonlinear problems successively and high computation rate are features that
are the reasons for many experts to prefer ANN. Moreover, there is no requirement for
any assumptions in ANN (Pal and Mitra, 1992). Neural networks can be classified
according to their architectures as single layer feed-forward networks, multilayer
feed-forward networks and recurrent neural network. The most important section of
ANN is training. Training is another name of seeking the correct weight values. While
classical statistical techniques only estimate the coefficient of independent variables,
ANN selects proper weights during training and keeps them for further use to predict
the output (Cheng and Titterington, 1994). There are three types of learning processes;
supervized, unsupervized and reinforcement learning.
2.4 Evolutionary computation
Evolutionary computation concept based on Darwin’s evolution theory by applying
the biological principle of natural evolution to artificial systems for the solution of
optimization problems has received significant attention during the last two decades,
Intelligence
decision systems
363
JEIM
24,4
364
although the origins were introduced in the late 1950s with works of Bremermann in
1962, Friedberg in 1958 and 1959, and Box in 1957, and others (Back et al., 1997). The
domain of evolutionary computation involves the study of the foundations and the
applications of computational techniques based on the principles of natural evolution.
Evolutionary algorithms employ this powerful design philosophy to find solutions to
hard problems from different domains, including optimization, automatic
programming, circuit design, machine learning, economics, ecology, and population
genetics, to mention but a few. Evolutionary computation for solving optimization and
other problems has considerable advantages. The first and important advantage is
adaptability for fluxional situations. Unlike many traditional optimization procedures
where the calculation must be restarted from the beginning if any variable in the
problem changes, evolutionary computation does not need to restart from beginning
since the current population serves as a memoir of stored knowledge that can be
applied on the fly to a dynamic environment. Therefore, traditional optimization
methods are more computationally expensive than the evolutionary computation.
Another advantage of an evolutionary computation to problem solving comes in being
able to generate good enough solutions quickly enough for them to be of use
(Kahraman et al., 2010). There are several techniques of evolutionary computations,
among which the best known ones are genetic algorithms, genetic programming,
evolution strategies, classifier systems and evolutionary programming; though
different in the specifics they are all based on the same general principles (Back et al.,
1997; Dimopoulos and Zalzala, 2000; Pena-Reyes and Sipper, 2000).
2.4.1 Genetic algorithms. Genetic algorithms (GAs) which were first developed by
Holland in 1975 are based on mechanics of natural selection and genetics to search
through decision space for optimal solutions. The metaphor underlying GAs is natural
selection. Genetic algorithm (GA) is a search technique used in computing to find exact
or approximate solutions to optimization and search problems. GAs can be categorized
as global search heuristics. GAs are a particular class of evolutionary algorithms (also
known as evolutionary computation) that use techniques inspired by evolutionary
biology such as inheritance, mutation, selection, and crossover (also called
recombination). GAs are stochastic search methods based on the genetic process of
biological organisms. Unlike conventional optimization methods, GAs maintain a set
of potential solutions (populations) in each generation. A GA is encoding the factors of
a problem by chromosomes, where each gene represents a feature of problem. In
evolution, the problem that each species faces is to search for beneficial adaptations to
the complicated and changing environment. In other words, each species has to change
its chromosome combination to survive in the living world. In GA, a string represents a
set of decisions (chromosome combination), that is a potential solution to a problem.
Each string is evaluated on its performance with respect to the fitness function
(objective function). The ones with better performance (fitness value) are more likely to
survive than the ones with worse performance. Then the genetic information is
exchanged between strings by crossover and perturbed by mutation. The result is a
new generation with (usually) better survival abilities. This process is repeated until
the strings in the new generation are identical, or certain termination conditions are
met. This algorithm is continued until the stopping criterion is reached. GAs are
different from other search procedures in the following ways:
.
.
.
GAs consider many points in the search space simultaneously, rather than a
single point;
GAs work directly with strings of characters representing the parameter set, not
the parameters themselves;
GAs use probabilistic rules to guide their search, not deterministic rules.
Because GAs consider many points in the search space simultaneously there is a
reduced chance of converging to local optima. In a conventional search, based on a
decision rule, a single point is considered and that is unreliable in multimodal space.
GAs consist of four main sections as: Encoding, Selection, Reproduction, and
Termination (Kaya and Engin, 2007; Engin et al., 2008; Kaya, 2009a, 2009b).
2.4.2 Genetic programming. Genetic programming (GP) developed by Koza (1990,
1992) is an evolutionary algorithm-based methodology inspired by biological evolution
to find computer programs that perform a user-defined task. It is defined by Koza
(1990) as “. . . genetic programming paradigm, populations of computer programs are
genetically bred using the Darwinian principle of survival of the fittest and using a
genetic crossover (recombination) operator appropriate for genetically mating
computer programs”. In GP, instead of encoding possible solutions to a problem as
a fixed-length character string, they are encoded as computer programs. The
individuals in the population are programs that – when executed – are the candidate
solutions to the problem. It is a specialization of genetic algorithms where each
individual is a computer program. Therefore, it is a machine learning technique used to
optimize a population of computer programs according to a fitness landscape
determined by a program’s ability to perform a given computational task (Kahraman
et al., 2010).
2.4.3 Evolution strategies. Evolution strategies use natural problem-dependent
representations, and primarily mutation and selection as search operators were
introduced by Ingo Rechenberg and Hans Paul Schweffel in the 1960s (Pena-Reyes and
Sipper, 2000) as a method for solving parameter-optimization problems. Mutation is the
major genetic operator in evolution strategies. It also plays the role of a reproduction
operator given that the mutated individual is viewed as an offspring for the selection
operator to work on. One of the commonly proposed advantages of evolution strategies
(ESs) is that they can be easily parallelized. ESs with l offspring per generation
(population size l) are usually parallelized by distributing the function evaluation for
each of the offspring on a different processor. However, when the number of offspring
is smaller than the number of available processors, the advantage of using evolution
strategies in parallel cannot be fully exploited. Consequently, when large numbers of
processors are available, it is desirable to develop an algorithm that can handle a large
population efficiently (Hansen et al., 2003).
2.4.4 Evolutionary programming. Evolutionary programming (EP) was first used
by Lawrence J. Fogel in the 1960s in order to use simulated evolution as a
learning process aiming to generate artificial intelligence. Fogel used finite state
machines as predictors and evolved them. It has been applied with success to
many numerical and combinatorial optimization problems in recent years. Like
with evolution strategies, evolutionary programming first generates offspring and
then selects the next generation. It is becoming harder to distinguish it from
evolutionary strategies. Some of its original variants are quite similar to the later
Intelligence
decision systems
365
JEIM
24,4
366
genetic programming, except that the program structure is fixed and its numerical
parameters are allowed to evolve. Formulating EP as a special case of the
generate-and-test method establishes a bridge between EP and other search
algorithms, such as evolution strategies, GAs, simulated annealing, tabu search,
and others, and thus facilitates cross-fertilization among different research areas
(Yao, 1999; Kahraman et al., 2010).
2.4.5 Classifier systems. Classifier systems (CSs), presented by J.H. Holland in the
1970s, are evolution-based learning systems, rather than a “pure” evolutionary
algorithm. They can be thought of as restricted versions of classical rule-based
systems, with the addition of input and output interfaces. A classifier system consists
of three main components:
(1) the rule and message system, which performs the inference and defines the
behavior of the whole system;
(2) the apportionment of credit system, which adapts the behavior by credit
assignment; and
(3) the GAs, which adapt the system’s knowledge by rule discovery (Pena-Reyes
and Sipper, 2000).
CSs exploit evolutionary computation and reinforcement learning to develop a set of
condition-action rules (i.e. the classifiers) which represent a target task that the system
has learned from on-line experience. There are many models of CSs and therefore also
many ways of defining what a learning classifier system is. Nevertheless, all CS
models, more or less, comprise four main components:
(1) a finite population of condition action rules, called classifiers, that represents
the current knowledge of the system;
(2) the performance component, which governs the interaction with the
environment;
(3) the reinforcement component, which distributes the reward received from the
environment to the classifiers accountable for the rewards obtained; and
(4) the discovery component, which is responsible for discovering better rules and
improving existing ones through a GA (Holmes et al., 2002).
2.4.6 Ant colony optimization. In the early 1990s, ant colony optimization (ACO) which
is a class of optimization algorithms modeled on the actions of a real ant colony was
introduced by M. Dorigo and colleagues as a novel nature-inspired metaheuristic for
the solution of hard combinatorial optimization problems. The inspiring source of ACO
is the foraging behavior of real ants (Dorigo and Gambardella, 1997). In general, the
ACO approach attempts to solve an optimization problem by repeating the following
two steps (Dorigo and Blum, 2005):
(1) Candidate solutions are constructed using a pheromone model, that is, a
parameterized probability distribution over the solution space.
(2) The candidate solutions are used to modify the pheromone values in a way that
is deemed to bias future sampling toward high quality solutions.
2.4.7 Particle swarm optimization. Particle swarm optimization (PSO) which is
population based stochastic optimization technique developed by Kennedy and
Eberhart (1995), inspired by social behavior of bird flocking or fish schooling is a
global optimization algorithm for dealing with problems in which a best solution can
be represented as a point or surface in an n-dimensional space. PSO shares many
similarities with evolutionary computation techniques such as genetic algorithms
(GA). The system is initialized with a population of random solutions and searches for
optima by updating generations. However, unlike GA, PSO has no evolution operators
such as crossover and mutation. In PSO, the potential solutions, called particles, fly
through the problem space by following the current optimum particles (Kahraman
et al., 2010).
Intelligence
decision systems
367
2.5 Hybrid systems
Although the sufficient results found by computational intelligence techniques, more
effective solutions can be obtained when used combination of these techniques. Each
combination has an aim to decrease the limitation of one method. For example, genetic
algorithm has been used to improve the performance of artificial neural networks in
literature. Usage of genetic algorithms in artificial neural networks for training or
finding the appropriate architecture can keep from getting trapped at local minima.
Another example is using fuzzy inference with other computational intelligence
techniques. A fuzzy inference system can take linguistic information (linguistic rules)
from human experts and also adapt itself using numerical data to achieve better
performance. Generally, using hybrid systems provides synergy to the resulting
system in the advantages of the constituent techniques and avoids their shortcomings
(Jang and Sun, 1995).
3. Literature review
In Table I, the frequency of the papers including the keywords “enterprise information
management (EIM)” and “intelligence techniques” are given for the years 2007-2011.
When a search was made with respect to the keywords “EIM” and “fuzzy set theory
(FST)” as an intelligence technique, totally 200 articles were viewed. When it was made
with respect to the keywords “EIM” and “intelligence techniques”, totally 531 articles
were viewed. At the same time, these papers were classified based on their keywords.
For example, the number of the papers using the “multi-agent systems (MAS)” and
“neural networks (NNs)” are 147 and 108, respectively.
A pie chart which summarizes the distribution of the papers on intelligence decision
systems in enterprise information management is given in Figure 1. As it is seen from
Year
K: EIM
K: FST
K: EIM
K: MAS
K: EIM
K: NNs
K: EIM
K: GA
K: EIM
K: ACO
K: EIM
K: PSO
2011
2010
2009
2008
2007
Total
2
43
67
57
31
200
1
34
35
45
32
147
2
26
34
28
18
108
1
10
18
17
6
52
6
2
2
1
11
7
2
3
1
13
Table I.
The results based on
publication years
JEIM
24,4
368
Figure 1.
Distribution of papers on
intelligence decision
systems in enterprise
information management
Figure 1, the papers have taken more attention from year to year. Of these papers,
25 percent have been published in the last two years. It can be concluded that
intelligence decision systems in enterprise information management will take more
and more attention in the next years.
The percentages of the published papers on intelligence techniques in EIM are
shown in Figure 2. As can be seen from this pie chart, the FST is the most used
technique for EIM with 38 percent whereas the MAS and NN are the successors with 28
percent and 20 percent, respectively. PSO and ACO are the least used two intelligence
techniques for this aim with 2 percent.
The papers related to the FST in EIM are also classified based on their subject areas
and the frequency distribution of these papers as in Figure 3. As it is seen from
Figure 3, the most of these papers are related to “computer science” and totally 155
papers have been published in this area. The next areas are “engineering” and
“decision sciences” with 154 and 61 papers, respectively.
Chang et al. (2011) list the criteria that influence supplier selection, and constructs
the strategy map among these criteria using the fuzzy set theory. The strategy map
revealed interdependencies among these criteria and their strengths. Chen et al. (2010)
Figure 2.
Percentages of the
published papers on
intelligence techniques in
EIM
Intelligence
decision systems
369
Figure 3.
Papers related to the FST
in EIM
introduced fuzzy trust evaluation method for sharing knowledge derived from the
activities of a virtual enterprise (VE) and the interactions among allied enterprises,
including collaborative relations and the period of collaboration, is developed.
Consisting of the evaluation of the correlation among activities, current trust and of
integral trust, the method assessed current trust via fuzzy rule based system having
the inputs of collaborative time, activity correlation and collaborative relation. The
study had a potential of not only enhancing information sharing, but also increasing
the efficiency and competitiveness of VEs. Chen and Lin (2009) proposed a fuzzy
linguistic performance index based on flow network model to evaluate the performance
of an ERP system depending upon the linguistic grades of an ERP examination of the
company users involved. The study contributed to ERP pre-implementing for making
decision. Tai and Chen (2009) proposed a suitable model for intellectual capital
performance evaluation by combining 2-tuple fuzzy linguistic approach with multiple
criteria decision-making (MCDM) method. Karsak and Ozogul (2009) used QFD, fuzzy
linear regression and 0 –1 goal programming by integrating them in the ERP selection
problem. ERP characteristics were obtained from vendors in the market and customer
requirements by regarding the company profile and strategic selection criteria. ERP
system selection was performed with respect to weighted sum of deviations from the
maximum achievable values for company needs. Xirogiannis et al. (2008) discussed a
novel approach of designing a decision-modeling tool, which assesses the impact of
contemporary human resource management (HRM) practices to the shareholder value
and satisfaction. The proposed methodology utilized the fuzzy causal characteristics of
fuzzy cognitive maps (FCMs) to generate a hierarchical and dynamic network of
interconnected HR performance drivers. The intelligent computing characteristics of
FCMs were also employed to establish a dynamic feedback and bi-directional
alignment of HRM practices and strategic improvement. Zhang and Lu (2007) proposed
a fuzzy bilevel decision making model for a general logistics planning problem and
develops a fuzzy number based Kth-best approach to find an optimal solution for the
proposed fuzzy bilevel decision problem. The proposed approach illustrated an optimal
solution in logistics management, which meets maximally/minimally the objectives of
JEIM
24,4
370
Figure 4.
A pie chart for papers
related to MAS for EIM
both supplier and distributor (or other parts of the logistics chain). Kulak et al. (2005)
employed information axiom to technology project selection problem. Return on
investment, user satisfaction, operational agility, and strategic competitiveness were
considered under both crisp and fuzzy environment. Liu et al. (2005) presented a
competence appraisement model of human resource in an enterprise. The paper put
forward the appraisement index architecture of enterprise human resource competence
on the foundation of adopting the concept and models of competence, and establishes a
general model of the enterprise human resource competence appraisement. Li et al.
(2003) developed an analytical formwork which captures the process of information
flowing, and then established a practical method on fuzzy set theory to assess
effectiveness of information flow control. With result of study, the enterprise could
conduce itself to have cognizance of enterprise information management
comprehensively. Harigopal (2001) proposed cognizant enterprise maturity model is
a unique five-level maturity model that provides a roadmap for
businesses/organizations to evolve their capabilities for realizing immediate value.
The application could be meaningful for standardization and stimulate more
cross-disciplinary research and development in cognizant enterprise. Chan et al. (2000)
proposed an integrated approach for the automatic design of flexible manufacturing
systems (FMS), which uses simulation and fuzzy multi-criteria decision-making
techniques. The design process consists of constructing and testing the alternative
designs using simulation methods. The selection of the most suitable design based on a
multi-criteria decision-making technique, the fuzzy analytic hierarchy process, was
employed to analyze the output from the FMS simulation models.
MAS are the second most used technique for EIM. The papers related to MAS for
EIM are classified based on their subject areas. A pie chart for this analysis is
illustrated in Figure 4. As it is seen from Figure 4, the area “Computer Science” is the
most popular area for the application of MAS with 41 percent whereas the areas
“Engineering” and “Decision Sciences” are the following areas with the percentages 38
and 9, respectively.
Liu et al. (2011) concerned with unexpected events (i.e. inventory failure, increased
demand fluctations, and delivery delay) in e-procurement by using multi-agent system.
Having a hierarchical framework, the system was simulated within a restaurant
e-procurement. The study of Fu et al. (2008) addressed the problems, such as sharing of
information, the ability to extend and re-engineer, and the reusable ability of legacy
systems in distributed and heterogeneous environments. A method based on agent and
ontology was proposed to design knowledge management system, which includes two
agencies, namely knowledge and application agencies. The effectiveness of the system
for small and medium-size enterprises was tested. Cerrada et al. (2007) described a
model for managing faults in industrial processes. The model had a generic framework
that uses multi agent system (MAS) for distributed control systems. The system
managed faults with feedback control process and decides about the scheduling of the
preventive maintenance tasks, also running preventive and corrective specific
maintenance tasks. Monteiro et al. (2007) addressed a hierarchical architecture to
integrate individual planner agent, negotiator agent, and mediator agent with a
decentralized control for achieving robustness and flexibility of the supply chain
network. The negotiator agent was responsible for limiting the negotiation process and
facilitating cooperation between production centers. Qiu et al. (2006) argued that
workflow management technology promotes the automation of enterprise business
processes. Multi-agents were utilized in the decomposed material flow management
instances that directly interact with enterprise workflow systems. That information on
material usage and movement can be reflected to planning department in a timely
manner was provided. Kishore et al. (2006) synthesized literature in the areas of
integrated business information systems (IBIS) systems and multi-agent systems with
the intent of developing comprehensive foundation ontology for the universe of
Multiagent-based Integrative Business Information Systems (MIBIS). Based on this
review and synthesis, they identified a set of eight ontological constructs (i.e. goal, role,
interaction, task, information, knowledge, resource, and agent) for MIBIS modeling
that are minimally required to model a system efficiently, precisely, and
unambiguously in the MIBIS universe. Caridi et al. (2006) conduct the analysis of
the multi-agent system to automate and optimize collaboration along the supply chain
via introducing two consecutive multi-agent systems. They focused on Collaborative
Planning Forecasting and Replenishment (CPFR), a nine-step approach which provide
volunteer standards, protocols, guidelines, etc. required to exchange sales and order
forecasts between trading partners. The system had the potential of including
inventory management, sales management and production management, along all the
supply chain. Providing a comprehensive review of studies about intelligent agents
applications for supply chain, Mangina and Vlachos (2005) focused on the effects of
MAS on beverage supply chains. It was argued that the efficiency might be increased
through the elimination of unexpected mistakes, information visibility and information
quality improves, excessive inventory and severe delays will be reduced, capacities
will be balanced and customer service will be improved by pre-processing and filtering
of information. While discussing hybrid multi-agent system architecture for enterprise
integration using a computer network, Nahm and Ishikawa (2005) described that the
discrete real-world was dynamic because of a diverse, frequently changing situation,
Intelligence
decision systems
371
JEIM
24,4
372
Figure 5.
A pie chart for papers
related to NNS for EIM
and there will be no obligation for a person or computerized agent to remain with a
network for a certain period. They proposed a hybrid multi-agent architecture for
enterprise that updates internal and external states regularly. In the research of Li et al.
(2004), an agent that incorporates the working mechanism of expert systems was
designed to assist a designer with the aim of shortening the product design cycle.
Symeonidis et al. (2003) proposed an intelligent policy recommendation multi-agent
system (that introduces adaptive intelligence as an add-on for ERP software
customization. They combined multi-agent and data mining technologies to provide
intelligent decision support to supply chain management. A flexible multi agent
system for supply chains that can adapt to transaction changes brought about by new
products or trading partners was presented in Ahn et al. (2003). The system
contributed to the flexibility of adapting to new products or new trading partners that
involve changes in transactions. Anumba et al. (2001) emphasized the potential of
multi-agents systems (MAS) for collaborate design the charactersitics of which were
interdisciplinary interaction and negotiation in design. Agent coordination issues for
decision support applications was addressed by Bui and Lee (1999) and taxonomy of
agent characteristics was proposed. A development lifecycle that looks at agent-based
DSS as being a design of coordinated agents to optimally support a problem-solving
process was also presented.
The other technique used in EIM is NNs. As it is seen from Figure 5, NNs have
generally been used in the area “Computer Science” for EIM and the percentage of this
area is 40. The second most popular area for the using of NNs is the area “Engineering”
with 29 percent.
Lee et al. (2011) focused on demand and supply chain management and examined
how artificial intelligence techniques and RFID technology could enhance the
responsiveness of the logistics workflow. The proposed system determined the correct
replenishment strategy by automatically classifying the distribution patterns within
the complex demand and supply chain. In a recent study, Zhao and Yu (2011)
introduced a case based reasoning framework for supplier selection in petroleum
enterprises. The method based on data mining techniques which solves three key
problems of case reasoning system, includes calculating the weights of the attributes
with information entropy in case warehouse organizing process objectively, evaluating
the similarities with k-prototype clustering between the original and target cases in
case retrieving process exactly, and extracting the potential rules with back
propagation neural networks from conclusions in maintenance and revising process
efficiently. Addressing how to coordinate a rule-based system and supervized learning
for making more flexible the production planning activity, López-Ortega and
Villar-Medina (2009) utilized a feed-forward neural network (FANN), which was
embedded in a machine agent with the aim of determining the appropriate machine in
order to fulfill clients’ requirements. Also, an expert system was provided to a tool
agent, which in turn was in charge of inferring the right tooling. Chen and Du (2009)
aimed at the financial and the non-financial ratios in the financial statement, and used
the back-propagation network (BPN), a neural network that used a supervized learning
method and feed-forward architecture, and the clustering model to compare the
performance of the financial distress predictions, in order to find a better early-warning
method. The authors concluded that the artificial neural network (ANN) approach
obtains better prediction accuracy than the data mining clustering approach by
developing a financial distress prediction model. Chang et al. (2008) proposed a neural
network evaluation model for Enterprise Resource planning (ERP) performance from
Supply Chain Management (SCM) perspective. The survey data were gathered from a
transnational textile firm in Taiwan. The authors indicated that while the external
environments and alliance partnerships facing an enterprise were becoming more
complex, executive should enhance efficiency and performance of supply chain
management as well as to gain potential competitive advantages. Considering price,
quality, reputation, service and improvement culture of suppliers, Choy et al. (2002)
used an artificial neural network to design an intelligent supplier relationship
management system in order to benchmark suppliers’ performance and shorten the
cycle time of outsourcing. Moreover, the model was applied for new product
development process in a company in Hong Kong.
The other intelligence techniques such as GA, ACO and PSO have been used in
EIM. In this subsection, some papers using GA, ACO and PSO in EIM in the literature
are briefly summarized. Meilin et al. (2010) addressed the problem of dynamic job shop
scheduling. Introducing a new ant moving strategy, the methodology utilized the
real-time feedback from the Manufacturing Execution System Intelligent Data
Terminal was in the new pheromone updating rule. Its application in the production
process in a mould and Die enterprise lead to 23 percent decrease in the manufacturing
cycle of products. In addition to the achievements about scheduling performance, the
methodology enhanced management control in workshops via reducing the workload
of the workshop manager by 80 percent. Irani et al. (2009) presented a model that
defines a relationship between knowledge management and organizational learning,
and highlights factors that can lead a firm to develop itself towards a learning
organization via fuzzy cognitive mapping. The study revealed that the model
developed showed that a relationship does exist between knowledge management and
organizational learning and each knowledge concept engenders/fosters realization of
the other. Xiaobing et al. (2008) proposed an ACO model for the minimization of costs
Intelligence
decision systems
373
JEIM
24,4
374
in alloy management. In the proposed model, the usage of each element was tried to be
kept close to the pre-determined target value. When compared to linear programming
model, the effectiveness of ACO model was demonstrated in terms of computational
burden and solution quality by means of the application in Enterprise Resource
Planning (ERP) integrated information system of a special steel enterprise. Sun and Li
(2007) developed a multi-classification model for the evaluation of a customer service
center of a power supply enterprise which adopts Customer Relationship Management
(CRM). Considering economic performance, social performance and business quality of
the system, the model utilizes principal component for assigning weights to indicators.
As an original feature, evaluation classes are determined via decision tree. The
evaluation result shows that the model has better accuracy of the classification and
practicality. In the research of classifying and segmentation with RFM (recency,
frequency and monetary) field, Cheng and Chen (2009) employed RFM attributes and
K-means algorithm into Rough set theory to mind accuracy classification rules which
can be used for driving an CRM in an enterprise. Within the model, after continous
attributes are discretized for rough set theory, costumers are clustured. Then, the
characteristic of customer is determined via extracting rules. The methodology
outperforms decision tree, neural network and naive bayes in terms of accuracy rate
regardless of the number of generated classess on output, and yields understandable
decision rules. Providing a classification of artificial and computational intelligence
techniques, Wu (2010) presented a business intelligence system including optimizer,
predictor and classifier. The author suggested that real-coded genetic algorithm (RGA)
or artificial immune system (AIS) should be used to optimize the parameter settings of
the intelligent business intelligence system. Kuo et al. (2010) introduced a genetic
programming-based knowledge-evolution framework to search for a good integrated
classification tree with different evolving time points. In the proposed model, each
member was represented as a classification tree (CT), and each CT represents a rule set,
which is randomly generated, completely by the identical random function and not
limited for original knowledge. Herrera et al. (2001) utilized a genetic algorithm that
considered the requirements of positions and the relationships among candidates, and
solved personnel assignment problem with verbal information. Both personnel skills
and position requirements are expressed via fuzzy numbers. Emphasizing the
importance of both geographical information systems and internet on bridge
maintenance management, Liu and Itoh (2001) applied genetic algorithm for the
rehabilitation plan of bridge decks with the aim of minimizing rehabilitation cost and
deterioration degree simultaneously. Lu et al. (2009, 2010) presented a novel risk
management model for Virtual Enterprise (VE), a Constructional Distributed Decision
Making (CDDM) model where the situation of information symmetry between owner
and partners is considered. They considered various risks for VE, due to VE’s agility
and diversity of its members and its distributed characteristics and proposed a
Multi-swarm Particle Swarm Optimization (MPSO) to solve the resulting optimization
problem. In the hybrid method of Gao et al. (2009) the wide range of mapping
capabilities of artificial neural network (ANN) and the global search ability of particle
swarm optimization were combined to find the global optimal solution during ANN
training process. The method was applied to evaluate business efficiency in terms of
financial statements. Niu and Gu (2007) established a hybrid genetic particle swarm
optimization algorithm for a material purchase and storage optimization for electric
power plants to minimize the cost based on its characteristic of raw material stock. The
algorithm combined the evolution idea of genetic algorithm (GA) with population
intellectual technique of particle swarm optimization (PSO) algorithm.
4. Conclusions
In this paper, the roles of intelligence techniques in EIM are discussed to obtain a
successful business strategy. EIM systems in many companies have been developed
following the needs arising from administration, control, reporting and transaction
management. EIM focus must be shifted from general management and control to a
development of the means and solutions to enable integration of cross-functional
teams, key business processes, performance management, information and knowledge
to increase profitable market share.
Intelligence techniques are rapidly emerging as new tools in information
management systems. Especially, intelligence techniques can be used to utilize
decision process of EIM. These techniques can increase sensitiveness, flexibility and
accuracy of information management systems. Creation and implementation of EIM
strategy and investment decisions needs to be guided by not only intelligence
techniques but also business strategy and needs. If it successes, the intelligence
decision systems arrive its best benefit.
The hybrid systems that contain two of more intelligence techniques will be more
used in the future. So, the hybrid intelligence systems in enterprise information
management should get the necessary importance from practitioners and
academicians.
References
Ahn, H.J., Lee, H. and Park, S.J. (2003), “A flexible agent system for change adaptation in supply
chains”, Expert Systems with Applications, Vol. 25, pp. 603-18.
Anumba, C.J., Ugwu, O.O., Newnham, L. and Thorpe, A. (2001), “A multi-agent system for
distributed collaborative design”, Logistics Information Management, Vol. 14 No. 5,
pp. 355-66.
Back, T., Hammel, U. and Schwefel, H.P. (1997), “Evolutionary computation: comments on the
history and current state”, IEEE Transactions on Evolutionary Computation, Vol. 1 No. 1,
pp. 3-17.
Baron, J. (1998), Thinking and Deciding, Cambridge University Press, Cambridge.
Bui, T. and Lee, J. (1999), “An agent-based framework for building decision support systems”,
Decision Support Systems, Vol. 25, pp. 225-37.
Caridi, M., Cigolini, R. and De Marco, D. (2006), “Linking autonomous agents to CPFR to improve
SCM”, Journal of Enterprise Information Management, Vol. 19 No. 5, pp. 465-82.
Cerrada, M., Cardillo, J., Aguilar, J. and Faneite, R. (2007), “Agents-based design for fault
management systems in industrial processes”, Computers in Industry, Vol. 58, pp. 313-28.
Chan, F.T.S., Jiang, B. and Tang, N.K.H. (2000), “The development of intelligent decision support
tools to aid the design of flexible manufacturing systems”, International Journal of
Production Economics, Vol. 65, pp. 73-84.
Chang, B., Chang, C.W. and Wu, C.H. (2011), “Fuzzy DEMATEL method for developing supplier
selection criteria”, Expert Systems with Applications, Vol. 38 No. 3, pp. 1850-8.
Intelligence
decision systems
375
JEIM
24,4
376
Chang, I.C., Hwang, H.G., Liaw, H.C., Hung, M.C., Chen, S.L. and Yen, D.C. (2008), “A neural
network evaluation model for ERP performance from SCM perspective to enhance
enterprise competitive advantage”, Expert Systems with Applications, Vol. 35, pp. 1809-16.
Chen, S.G. and Lin, Y.K. (2009), “On performance evaluation of ERP systems with fuzzy
mathematics”, Expert Systems with Applications, Vol. 36 No. 3, pp. 6362-7.
Chen, T.Y., Chen, Y.M., Lin, C.J. and Chen, P.Y. (2010), “A fuzzy trust evaluation method for
knowledge sharing in virtual enterprises”, Computers & Industrial Engineering, Vol. 59
No. 4, pp. 853-64.
Chen, W.S. and Du, Y.K. (2009), “Using neural networks and data mining techniques for the
financial distress prediction model”, Expert Systems with Applications, Vol. 36, pp. 4075-86.
Cheng, B. and Titterington, D.M. (1994), “Neural networks: a review from a statistical
perspective”, Statistical Science, Vol. 9 No. 1, pp. 2-54.
Cheng, C.H. and Chen, Y.S. (2009), “Classifying the segmentation of customer value via RFM
model and RS theory”, Expert Systems with Applications, Vol. 36 No. 3, pp. 4176-84.
Choy, K.L., Lee, W.B. and Lo, V. (2002), “An intelligent supplier management tool for
benchmarking suppliers in outsource manufacturing”, Expert Systems with Applications,
Vol. 22, pp. 213-24.
Dimopoulos, C. and Zalzala, A.M.S. (2000), “Recent developments in evolutionary computation
for manufacturing optimization: problems, solutions, and comparisons”, IEEE
Transactions on Evolutionary Computation, Vol. 4 No. 2, pp. 93-113.
Dorigo, M. and Blum, C. (2005), “Ant colony optimization theory: a survey”, Theoretical
Computer Science, Vol. 344, pp. 243-78.
Dorigo, M. and Gambardella, L.M. (1997), “Ant colonies for the travelling salesman problem”,
BioSystems, Vol. 43, pp. 73-81.
Engin, O., Çelik, A. and Kaya, İ. (2008), “A fuzzy approach to define sample size for attributes
control chart in multistage processes: an application in engine valve manufacturing
process”, Applied Soft Computing, Vol. 8 No. 4, pp. 1654-63.
Fu, L. (1994), Neural Networks in Computer Intelligence, McGraw-Hill, New York, NY.
Fu, R.X., Yue, X., Song, M. and Xin, Z.H. (2008), “An architecture of knowledge management
system based on agent and ontology”, The Journal of China Universities of Posts and
Telecommunications, Vol. 15 No. 4, pp. 126-30.
Gao, Y., Gu, Y. and Li, T. (2009), “Evaluation approach on enterprise integrated business
efficiency based on ANN-QPSO”, 2009 International Conference on Information
Management, Innovation Management and Industrial Engineering, pp. 371-4.
Hansen, N., Müller, S.D. and Koumoutsakos, P. (2003), “Reducing the time complexity of the
derandomized evolution strategy with covariance matrix adaptation (CMA-ES)”,
Evolutionary Computation, Vol. 11 No. 1, pp. 1-18.
Harigopal, U. (2001), “Cognizant Enterprise Maturity Model (CEMM)”, IEEE Transactions on
Systems, Man, and Cybernetıcs – Part C: Applications and Reviews, Vol. 31 No. 4,
pp. 449-59.
Herrera, F., Lopez, E., Mendana, C. and Rodriguez, M. (2001), “A linguistic decision model for
personnel management solved with a linguistic biobjective genetic algorithm”, Fuzzy Sets
and Systems, Vol. 118 No. 1, pp. 47-64.
Holmes, J.H., Lanzi, P.L., Stolzmann, W. and Wilson, S.W. (2002), “Learning classifier systems:
new models, successful applications”, Information Processing Letters, Vol. 82, pp. 23-30.
Holsapple, C. and Whinston, A. (1987), Business Expert Systems, McGraw-Hill, New York, NY.
Irani, Z., Sharif, A.M. and Love, P.E.D. (2009), “Mapping knowledge management and
organizational learning in support of organizational memory”, International Journal of
Production Economics, Vol. 122 No. 1, pp. 200-15.
Jackson, P. (1998), Introduction to Expert Systems, Addison Wesley, Boston, MA.
Jahanshahloo, G.R., Lotfi, F.H. and Izadikhah, M. (2006), “Extension of the TOPSIS method for
decision-making problems with fuzzy data”, Applied Mathematics and Computation,
Vol. 181, pp. 1544-51.
Jang, J.S.R. and Sun, C.T. (1995), “Neuro-fuzzy modeling and control”, Proceedings of the IEEE,
Vol. 83 No. 3, pp. 378-406.
Kahraman, C., Kaya, İ. and Çınar, D. (2010), “Computational intelligence: past, today and future”,
in Ruan, D. (Ed.), Computational Intelligence in Complex Decision Systems, Atlantis Press,
Paris, pp. 1-46.
Karsak, E.E. and Ozogul, C.O. (2009), “An integrated decision making approach for ERP system
selection”, Expert Systems with Applications, Vol. 36 No. 1, pp. 660-7.
Kaya, İ. (2009a), “A genetic algorithm approach to determine the sample size for attribute control
charts”, Information Sciences, Vol. 179 No. 10, pp. 1552-66.
Kaya, İ. (2009b), “A genetic algorithm approach to determine the sample size for control charts
with variables and attributes”, Expert Systems with Applications, Vol. 36 No. 5, pp. 8719-34.
Kaya, İ. and Çınar, D. (2008), “Facility location selection using a fuzzy outranking method”,
Journal of Multiple-Valued Logic and Soft Computing, Vol. 14, pp. 251-63.
Kaya, İ. and Engin, O. (2007), “A new approach to define sample size at attributes control chart in
multistage processes: an application in engine piston manufacturing process”, Journal of
Materials Processing Technology, Vol. 183 No. 1, pp. 38-48.
Kennedy, J. and Eberhart, R.C. (1995), “Particle swarm optimization”, Proceedings of the IEEE
International Conference on Neural Networks, pp. 1942-8.
Kishore, R., Zhang, H. and Ramesh, R. (2006), “Enterprise integration using the agent paradigm:
foundations of multi-agent-based integrative business information systems”, Decision
Support Systems, Vol. 42, pp. 48-78.
Koza, J.R. (1990), “Genetic programming: a paradigm for genetically breeding populations of
computer programs to solve problems”, Technical Report STANCS-90-1314, Stanford
University, Stanford, CA.
Koza, J.R. (1992), Genetic Programming, MIT Press, Cambridge, MA.
Kulak, O., Kahraman, C., Öztaysi, B. and Tanyas, M. (2005), “Multi-attribute information
technology project selection using fuzzy axiomation”, Journal of Enterprise Information
Management, Vol. 18 No. 3, pp. 275-88.
Lee, C.K.M., Ho, W., Ho, G.T.S. and Lau, H.C.W. (2011), “Design and development of logistics
workflow systems for demand management with RFID”, Expert Systems with
Applications, Vol. 38, pp. 5428-37.
Li, Y.J., Li, J.-F., Qi, W., Wang, H., Xie, A.S. and Jin, S.W. (2003), “Research on fuzzy assessment of
effectiveness for information flow control”, International Conference on Machine Learning
and Cybernetics, Vol. 4, pp. 2587-90.
Li, Y.L., Shao, X.Y., Li, P.G. and Liu, Q. (2004), “Design and implementation of a process-oriented
intelligent collaborative product design system”, Computers in Industry, Vol. 53, pp. 205-29.
Liu, C. and Itoh, Y. (2001), “Information technology applications for bridge maintenance
management”, Logistics Information Management, Vol. 14 Nos 5-6, pp. 393-400.
Intelligence
decision systems
377
JEIM
24,4
378
Liu, Q., Sun, S.X., Wang, H. and Zhao, J. (2011), “A multi-agent based system for e-procurement
exception management”, Knowledge-Based Systems, Vol. 24, pp. 49-57.
Liu, X., Ruan, D. and Xu, Y. (2005), “A study of enterprise human resource competence
appraisement”, Journal of Enterprise Information Management, Vol. 18 No. 3, pp. 289-315.
López-Ortega, O. and Villar-Medina, I. (2009), “A multi-agent system to construct production
orders by employing an expert system and a neural network”, Expert Systems with
Applications, Vol. 36 No. 2, pp. 2937-46.
Lu, F.-Q., Huang, M. and Wang, X.-W. (2010), “Partners’ risk level considered CDDM model for
risk management of Virtual Enterprise”, 2010 International Conference on Management
and Service Science, MASS 2010, pp. 1-4.
Lu, F.-Q., Huang, M., Ching, W.K., Wang, X.W. and Sun, X.L. (2009), “Multi-swarm particle
swarm optimization based risk management model for virtual enterprise”, Proceedings of
the 1st ACM/SIGEVO Summit on Genetic and Evolutionary Computation, pp. 387-92.
Mangina, E. and Vlachos, I.P. (2005), “The changing role of information technology in food and
beverage logistics management: beverage network optimization using intelligent agent
technology”, Journal of Food Engineering, Vol. 70, pp. 403-20.
Meilin, W., Xiangwei, Z., Qingyun, D. and Jinbin, H. (2010), “A dynamic schedule methodology
for discrete job shop problem based on ant colony optimization”, ICIME 2010 – 2nd IEEE
International Conference on Information Management and Engineering, Vol. 5, pp. 306-9.
Monteiro, T., Daniel, R.B. and Anciaux, D. (2007), “Multi-site coordination using a multi-agent
system”, Computers in Industry, Vol. 58, pp. 367-77.
Nahm, Y.E. and Ishikawa, H. (2005), “A hybrid multi-agent system architecture for enterprise
integration using computer networks”, Robotics and Computer-Integrated Manufacturing,
Vol. 21, pp. 217-34.
Niu, D. and Gu, X. (2007), “Application of HGPSOA in electric power system material purchase
and storage optimization”, Proceedings – ICSSSM’07: 2007 International Conference on
Service Systems and Service Management, pp. 1-7.
Pal, S.K. and Mitra, S. (1992), “Multilayer perception, fuzzy sets, and classification”, IEEE
Transactions on Neural Networks, Vol. 3 No. 5, pp. 683-97.
Pena-Reyes, C.A. and Sipper, M. (2000), “Evolutionary computation in medicine: an overview”,
Artificial Intelligence in Medicine, Vol. 19, pp. 1-23.
Qiu, R.G., Tang, Y. and Xu, Q. (2006), “Integration design of material flow management in an
e-business manufacturing environment”, Decision Support Systems, Vol. 42 No. 2,
pp. 1104-15.
Sugumaran, V. (2007), Application of Agents and Intelligent Information Technologies (Advances
in Intelligent Information Technologies), IGI Publishing, Hershey, PA.
Sun, W. and Li, S. (2007), “PCA-SVM-based comprehensive evaluation for customer relationship
management system of power supply enterprise”, Proceedings of the Sixth International
Conference on Machine Learning and Cybernetics, Hong Kong, 19-22 August 2007,
pp. 3811-4.
Symeonidis, A.L., Kehagias, D.D. and Mitkas, P.A. (2003), “Intelligent policy recommendations
on enterprise resource planning by the use of agent technology and data mining
techniques”, Expert Systems with Applications, Vol. 25 No. 4, pp. 589-602.
Tai, W.S. and Chen, C.T. (2009), “A new evaluation model for intellectual capital based on
computing with linguistic variable”, Expert Systems with Applications, Vol. 36, pp. 3483-8.
Turban, E., Volonio, L., McLean, E. and Wetherbe, J. (2009), Information Technology for
Management: Transforming Organizations in the Digital Economy, Wiley, New York, NY.
Wu, J.-Y. (2010), “Computational intelligence-based intelligent business intelligence system:
concept and framework”, 2nd International Conference on Computer and Network
Technology, ICCNT 2010, pp. 334-8.
Xiaobing, L., Ruilin, P., Qiunan, M. and Fajing, C. (2008), “Study on recipe cost optimization
system based on ant colony algorithms”, International Seminar on Business and
Information Management, Vol. 2008, pp. 287-90.
Xirogiannis, G., Chytas, P., Glykas, M. and Valiris, G. (2008), “Intelligent impact assessment of
HRM to the shareholder value”, Expert Systems with Applications, Vol. 35, pp. 2017-31.
Yao, X. (1999), “Evolutionary programming made faster”, IEEE Transactions on Evolutionary
Computation, Vol. 3 No. 2, pp. 82-102.
Zadeh, L.A. (1975), “The concept of a linguistic variable and its application to approximate
reasoning”, Information Sciences, Vol. 8, pp. 199-249.
Zadeh, L.A. (1996), “Fuzzy logic equals computing with words”, IEEE Transactions on Fuzzy
Systems, Vol. 4 No. 2, pp. 103-11.
Zhang, G. and Lu, J. (2007), “Model and approach of fuzzy bilevel decision making for logistics
planning problem”, Journal of Enterprise Information Management, Vol. 20 No. 2,
pp. 178-97.
Zhao, K. and Yu, X. (2011), “A case based reasoning approach on supplier selection in petroleum
enterprises”, Expert Systems with Applications.
About the authors
Cengiz Kahraman has BSc, MSc, and PhD degrees in Industrial Engineering from Istanbul
Technical University. His main research areas are engineering economics, quality management
and control, statistical decision making, and fuzzy sets applications. He has published over 100
international journal papers and four edited books from Springer. He guest-edited many special
issues of international journals. He is presently the head of Industrial Engineering Department of
Istanbul Technical University.
İhsan Kaya received the BS and MSc degrees in Industrial Engineering from Selçuk
University. He also received PhD degree from Istanbul Technical University on Industrial
Engineering. Dr Kaya is currently an Assistant Professor Dr at Yıldız Technical University
Department of Industrial Engineering. His main research areas are process capability analysis,
quality management and control, statistical decision making, and fuzzy sets applications. İhsan
Kaya is the corresponding author and can be contacted at: [email protected]
Emre Çevikcan received the BS degree in Industrial Engineering from Yıldız Technical
University, the MSc degree and PhD degree in 2010 in Industrial Engineering from Istanbul
Technical University. He studied the scheduling of production systems for his PhD dissertation.
Emre Çevikcan is currently a lecturer in the Industrial Engineering Department of Istanbul
Technical University. His research has so far focused on the design of production systems
(assembly lines, production cells, etc.), lean production, scheduling, decision making and fuzzy
logic.
To purchase reprints of this article please e-mail: [email protected]
Or visit our web site for further details: www.emeraldinsight.com/reprints
Intelligence
decision systems
379