Integration of Decision Support Systems and Data Mining
for Improved Decision Making
Omar al-Ketbi and Marc Conrad
Institute for Research in Applicable Computing, University of Bedfordshire, Luton, U.K.
Keywords: Data Mining, Decision Support Systems, System Integration, Abu Dhabi Police.
Abstract: A data mining (DM) integrated decision support system (DSS) is suggested to improve the performance of
the DSS implemented in a police organisation. A prototype of the suggested system is provided. The paper
provides an insight into DM-DSS integrated systems in the literature, and uses the results of the
investigation as a basis for a suggested system tailored against the particularity of Abu Dhabi (AD) Police
Organisation in the United Arab Emirates. It is believed that in general DM processes are demanding in
terms of time and other resources. Hence, it is suggested that the process be reduced in order to
accommodate to the particularity of AD Police in terms of size and current system performance. The system
is perceived on this basis.
1 INTRODUCTION
Abu Dhabi (AD) Police operate with other UAE
police departments through the Ministry of Interior
to achieve a safer society. The Police serve four
major UAE districts: Abu Dhabi, Al-Ain City, the
external region, and the western region. AD Police
have several units, which include police patrol,
emergency response, crime investigation, and traffic.
The primary objective of the Police is to become an
intelligence-led, proactive police force that reacts to
the needs of society with the highest level of
integrity and training.
For this aim, AD Police have constantly
undergone development of their information systems
in order to integrate a range of processes that include
the human side as well as the software and hardware
deployment with the ultimate goal to improve the
accuracy and quality of their undertaken decisions.
AD Police have established the “Decision-Making
Support Center” to help the Police explore the future
challenges rather than just conduct research on the
current phenomena. The Center also helps in quality
assessment and control. In 2011, the Police
implemented a GIS (Geographic Information
System) in order to integrate people and processes
for making better decisions.
However, the outcomes have not been very
promising: slow and inefficient results have been
noticed compared to the set expectations. Decision
making is supported by the Quality Department that
serves as a hub which provides quantitative and
qualitative data received from different departments
and organisational bodies countrywide. This system
then would serve to support accurate and relevant
decisions. The decision process is usually not
straightforward and takes several factors into
consideration. Typical challenges are associated
with data formats, content, validity and reliability. It
is also important to acknowledge the human factor
involved in the decision making process. According
to Hofstede (2001) the Arab culture is characterised
by collectivism and high-power distance.
Collectivism and high-power distance are two
dimensions set by Hofstede in his attempt to
establish quantifiable aspects of culture.
Collectivism refers to a higher extent to which the
group is prioritised over the self-interest as opposed
to individualism. Hofstede asserts that cultures with
high power distance, such as Arab culture, tend to be
more collectivist. This has implications into how
training and raising awareness are implemented to
lead to effective methods of system development.
Abu Dhabi Police aim to realise the full and
potential value of the data that it acquires. By the
right framework and strategy, data that are highly
utilised will eventually improve quality.
Implementing data mining and combining currently
separate collections can provide better information
and hence knowledge to improve quality. A data
482
al-Ketbi O. and Conrad M..
Integration of Decision Support Systems and Data Mining for Improved Decision Making.
DOI: 10.5220/0004450604820489
In Proceedings of the 15th International Conference on Enterprise Information Systems (ICEIS-2013), pages 482-489
ISBN: 978-989-8565-59-4
Copyright
c
2013 SCITEPRESS (Science and Technology Publications, Lda.)
quality strategy will achieve consistent direction
towards optimal decision making in the organisation.
The improved data quality will then ensure that the
organisation is more able to make informed and
accurate decisions on policies and strategies.
Decision Support Systems (DSS’s) comprise
different aspects of software, hardware, and data, as
well as human inputs in order to help decision
makers improve and enhance their decisions based
on analytical processing of the available
information. DSS’s have found uses in different
types of organisations and wherever strategic
decisions are to be made with high uncertainty
involved. On the other hand, Data Mining (DM)
allows robust analysis of huge amounts of data in
order to discover useful relationships or patterns
based on advanced statistical methods.
This paper provides insights into a prototype of
an integrated DM-DSS (Data Mining – Decision
Support System) solution for AD Police to assist
decision makers in the decision making process
based on better knowledge of the stored data. The
solution can be implemented by integrating a set of
DM techniques in the already established DSS in the
organisation. The new system is expected to increase
the efficiency sought from the DSS by acquiring
knowledge of the data fed to the system.
2 BACKGROUND AND RELATED
WORK
Public organisations face real challenges of using the
correct analysis of huge amounts of data. These data
are used for producing statistical analyses and
forecasts on economic, social, health and education
issues, which are highly related to government
planning in aspects such as economic growth,
development of interest rates and inflation,
household income, education standards, crime trends
and climate change are a major input.
Decision support systems (DSS’s) are widely
used in businesses around the world for the main
aim of helping executives to make better decisions
based on advanced levels of data refined and
presented to them. According to Hardin and Chhieng
(2007) DSS’s refer to a class of computer-based
systems that help in the process of decision making.
Similarly Liu et al. (2010) report that a DSS is
commonly defined in the literature as an interactive
computer information system designed to support
solutions to problems with taking decisions.
Padhy et al., (2012) argue that the value of
strategic information systems is easily recognised
yet efficiency and speed are not the only factors of
competitiveness. The large amounts of data have
called for new methods to analyse and understand
the relationship in between this data. Conclusions
and inferences from these data need special tools
and techniques that are able to delve deeper than
traditional decision support systems can. Moreover,
the rapid development of data digitisation produced
data stored in organisations’ data warehouses, which
required efficient exploitation and knowledge
extraction. Consequently, traditional problem
solving DSS’s became less efficient and started to
decline in the 1990s (Liu et al., 2010). Liu et al.
(2010) classify the main challenges facing DSS’s in
supporting decision making. These challenges
include:
Changes in technology from database to data
warehouse and on-line analysis processing
(OLAP), from mainframe to client/server
architecture, and from single user model to
World Wide Web access;
Increasing business interconnections in a more
dynamic business environment and intelligence.
For this a variety of other information systems
have been proposed, such as supply chain
management (SCM), enterprise resource
planning (ERP), and customer relationship
management (CRM);
The continuous increase in complexity in
decision making which requires executives to
consider a vast number of inputs and a
considerable amount of knowledge.
As Han and Kamber (2006 pp.4-5) point out,
organisations are usually data rich but information
poor. Data mining techniques help to analyse data
and uncover important data patterns, which may
conrtribute to better, knowledge-based strategies. In
doing so, data mining helps to bridge the gap
between data and information, which usually
prevents realising knowledge.
Given the uses and nature of solutions based on
data mining and those based on decision support
systems, it can be confidently suggested that such
solutions can be integrated in order to offer optimal
solutions to knowledge-based decision making
processes. However, only few examples of using
data mining to support management decisions are
found in the literature which will be discussed in the
following.
For example, Abu-Naser et al., (2011) suggest a
DSS based on data mining techniques for optimising
e-Learning in educational institutions. The suggested
system was not implemented, but according to the
IntegrationofDecisionSupportSystemsandDataMiningforImprovedDecisionMaking
483
authors, integrating data mining functionality into a
single DSS will be promising. The authors believe
that such a system will enable the educational
institutions to realise the importance of the DSS-
produced information in optimising their adopted
learning strategies.
El Seddawy et al., (2012) propose a DM-based
DSS to support top level management to make a
good decision in any time. According to the authors,
the proposed system can help decision makers in the
banking sector to address decisions related to new
investments.
In the application area of health services, Kumar
et al., (2011) suggest the use of DM based on
decision tree algorithms to classify certain diseases
and compare the effectiveness and correction rate
among them in order to support decisions on the
diagnostic process. According to Kumar et al.,
(2011), traditional decision support systems
developed to assist physicians in the diagnostic
process often are based on static data which may be
out of date. Hence, a decision support system which
can learn the relationships between certain
parameters would be useful to physicians and
hospitals.
Mohemad et al., (2010) argue that traditional
support systems that are widely used in the
construction industry are not optimal. Despite efforts
to integrate and transform the whole construction
tendering processes into electronic or digital forms,
the use of unstructured documents either in hard
copy or digital are still widely present. The authors
stress the need to extract and represent information
in machine-readable formats, attained by integrating
data mining in DSS model, which they believe to be
a promising approach.
Liu et al., (2010) have conducted research into
integrated decision support systems (IDSS)
including DM agent-enhanced integrated DSS to
improve decision support performance. The
researchers conclude that the main challenges in
developing an integrated DSS are the trade-off
between loose and tight integration strategies within
the integration frameworks and the seamless
integration across data, models and processes within
the integration frameworks.
Srinivasan et al., (2011) also suggest using DM
intelligent agents in DSS for achieving higher work
efficiency. They suggest that such a system can
provide autonomy, mobility, and collaboration of
different agents in order to provide a simple and fast
solution. Srinivasan et al. maintain that using agents
as data mining techniques, can help decision makers
by providing a more robust and quick DSS in
resolving issues in any complex situation.
Mladenić et al., (2003) maintain that there has
been no systematic attempt to integrate DM and
DSS. Reasons behind that are many but mainly
include the nature of data mining processes that
combine computer science and statistics, which
create some confusion on what implementation
aspects may be suitable for managerial decisions.
The following table (Table 1) summarises the
different approaches to linking DM to DSS:
Table 1: Approaches to DM-DSS found in the literature.
Author Context Approach
Abu-Naser et al.,
(2011)
E-Learning in
educational
institutions
DM-based DSS for
optimised results
El Seddawy et al.,
(2012)
The banking
sector
DM-based DSS to
support top level
management to
make decisions in
any time
Kumar et al.,
(2011)
Disease
classification
DSS that can learn
the relationships
between certain
parameters
Srinivasan et al.,
(2011)
Higher work
efficiency
DM intelligent
agents in DSS
Liu et al., (2010)
Research on
different DSS’s
Integrated Decision
Support Systems
(IDSS) and DM
agent-enhanced
IDSS to improve
decision support
performance
Mohemad et al.,
(2010)
Construction
industry
Optimising DSS in
the construction
industry through DM
In an initiative to address the above challenges,
the EU sponsored from 1999 over a 39 month period
the SolEuNet project, which comprised a network of
expert teams from business and academia to meet
client's Data Mining and Decision Support needs
(Mladenić, 2001). The outcomes of the project were
promising. The project identified the main objectives
to improve collaboration and communication,
promote awareness of organisational resources and
achievements, and enable organisational learning
and dissemination of such knowledge. However,
certain difficulties were encountered as detailed
below.
According to the final report at the project
closure, the project team maintained that collecting
and managing knowledge is a very hard task that can
never be fully accomplished. The project team
considered that the set of knowledge management
tools to support e-collaboration in data mining and
decision support can be considerably extended,
improved and further integrated. The team suggested
ICEIS2013-15thInternationalConferenceonEnterpriseInformationSystems
484
including in the near future, database and data
transformation services, automatically built activity
logs for data mining and decision support, a data
mining advisor that, given a dataset, suggests data
mining algorithms, and a central model evaluation
service. Another important improvement suggested
by the team is to adopt a standardised description of
the knowledge produced in all phases, which would
significantly simplify communication among
distributed cooperating groups.
3 PROBLEM DEFINITION
The aim of retaining information is not anymore a
significant concern of the organisation as it is
common that a numerous number of transactions are
stored in its warehouses regularly; with only a few
of real relevance to the organisation’s decision
making process. The amount of information stored
in databases increases daily and goes beyond the
technical skills and human capacity to interpret that
information
Database management systems have advanced at
a faster rate than the techniques used for extracting
and utilising information to be used in making
decisions (Power, 2008), in the sense of using the
trend of the past endeavours to anticipate the future
tendency (Lv and Li, 2009). Obtaining, storing and
managing information in larger organisations are
now ordinary operations of business companies and
usually performed automatically by electronic data
repositories (Saxena and Rajpoot, 2009).
One of the efficient techniques used for this aim
(i.e. extracting useful information) is data mining.
Some of the organisation’s information may be in a
textual format described in natural languages or does
not have a structure like that present in data tables
and structured relational databases. This type of
information, found mainly in the form of electronic
documents and emails, particularly in organisations
with limited affordability (technical and financial),
cannot be used with traditional data mining tools,
and thus minimises their potential. The data miner
thus has to prepare the data for mining. This
operation, called pre-processing, is complex and
may take several months to complete, depending on
the size of the project, and require a significant part
of the organisational resources, and usually
performed only by large enterprises (Calderon et al.,
2003). Artificial Intelligence systems such as Neural
Networks and Fuzzy Inference algorithms might be
appropriate for finding the solution to this problem.
Law enforcement organisations can highly
benefit from a DM-DSS integrated system regarding
the vast and loosely related information these
organisations deal with. According to Oatley et al.,
(2006), computer technologies that support criminal
investigations are wide ranging and include
geographical information systems displays,
clustering and link analysis algorithms and the more
complex use of data mining technology for profiling
crimes or offenders and matching and predicting
crimes. They also argue that knowledge from
disciplines such as forensic psychology, criminology
and statistics are essential to the efficient design of
operationally valid systems. According to McCue
(2003) however, one of the biggest challenges in
using data mining and predictive analytics in law
enforcement is that most, if not all, data encountered
was never intended to be analysed. Therefore,
significant challenges associated with data form,
content, reliability, and validity must be constantly
evaluated and addressed.
The above issue applies as well to AD Police.
Given the state-of-the-art DSS implemented in the
organisation, little success has been seen in
improving the day-by-day tasks. The decision
process in the organisation is usually not
straightforward and involves many sub-processes.
Some of the challenges faced are related to data
formats, content, validity and reliability.
4 DATA MINING TECHNIQUES
Recent developments in Information Systems as
well as the availability of extensive business data
repositories and database management systems,
accompanied by the advances in computer systems
and algorithms, have provided a gateway to pattern
matching and data models (Hand et al., 2001).
Data Mining consists of a set of techniques
inferred from Statistics and Artificial Intelligence
with the specific aim of discovering new, useful,
relevant and non-trivial knowledge that may be
hidden in a large mass of data (Markov and Larose,
2007). There have been numerous examples of its
uses in areas such as Marketing, Economics,
Engineering, Medicine and others. These techniques
are briefly described below.
Several techniques mentioned in the literature of
data mining such as classification, neural network,
genetic algorithm and others have long been known
(Segall and Zhang, 2006). Statistics on the
contribution of data mining techniques to selecting
data and evaluating the results of data mining like
clustering are also usually considered. What
IntegrationofDecisionSupportSystemsandDataMiningforImprovedDecisionMaking
485
distinguishes the perception of data mining is the
development of techniques for data mining
applications on a large scale databases. These
techniques are generally applied to the data on a
small scale. In addition, several techniques from the
fields of databases for data transformation are also
an integral part of the process of data mining.
Depending on the application domain and user
interest, various types of techniques can be
identified and applied. Some of these techniques are
briefed below in the order of relevance to the DSS of
AD Police. For example, dimensionality reduction is
useful when the number of involved variables
exceeds the capacity of the DSS to perform better,
whereas modelling helps improve the DSS
operations by feeding it with a stripped-down
version of a collection of data which models the
entire collection.
4.1 Dimensionality Reduction
Dimensionality reduction is a mathematical
technique used to reduce the number of random
variables involved in a dataset. It uses projection
from one vector space onto another one of lower
dimension.
4.2 Correlation
Correlation is a statistical method used to depict
relationships between random variables through
analysing potential links and inferring the degree of
connectivity. Given the large number of loose data
available for the DSS, correlation helps establish
useful relationships among seemingly unrelated
variable.
4.3 Modelling
Modelling or mathematical modelling is a
description of observed behaviour, simplified by
ignoring certain details to simulate the behaviour of
a phenomenon. Modelling allows complex systems
to be understood and their behaviour to be predicted
within the scope of the generated model.
4.4 Association
Finding association is a data mining technique that
allows searching for simultaneously occurring items
that in the transactions database. Algorithms such as
DHP (Dual Heuristic Programming), GSP
(Generalised Sequential Pattern) and Apriori (Wang
et al, 2006) among others are examples of tools that
implement the task of discovery of association.
4.5 Classification
Classification is a method that consists of defining a
mathematical function that maps a set of records to
one another in a predefined set of categorical labels,
called classes. This function is applied to predict the
class new records fall under.
4.6 Regression
Regression includes a search for a function that
maps the records from a database to actual values.
This method is similar to classification, but being
restricted only for numeric values.
4.7 Clustering
Clustering is used to separate records in a database
into subsets or clusters, such that the elements of a
cluster share common properties that distinguish
them from other clusters. The objective of this task
is to maximise intra-cluster similarity and minimise
inter-cluster similarity. Unlike classification task,
which has predefined labels, clustering needs to
automatically identify the data groups to which the
user must assign labels. Some algorithms used for
implementing this method are K-Modes, K-means,
K -Prototypes, K-Medoids, among others. (Han et al,
2006).
4.8 Summarisation
This task is very common in KDD (Knowledge
Discovery in Databases), is to seek, identify and
indicate common features among data sets. Inductive
logic and genetic algorithms are some examples of
technologies that can be applied in summarisation.
4.9 Detection of Deviations
This technique helps to identify records in the
database whose characteristics do not meet the
normal standards. Statistics is the main resource
provider used by this technique.
4.10 Discovery Sequence
It is an extension of the technique of finding
associations that are sought frequent by considering
several transactions occurring over a period. The
technique of association can be adjusted to engage
the generalised mining association rules. The post-
ICEIS2013-15thInternationalConferenceonEnterpriseInformationSystems
486
processing step includes processing the knowledge
gained in data mining. Among the main tasks of the
post-processing step are: preparation and
organisation and may include the simplification of
charts, diagrams or reports demonstrating, in
addition to the conversion of the representation of
knowledge gained.
5 APPLICATIONS IN SCIENCE
AND BUSINESS
Lessons learned from science and business
applications of data mining can be transferred to a
certain degree to the situation of AD Police.
Practitioners have devised methods for obtaining
relevant information from a large set of data, such as
dimensionality reduction, correlations, modelling
and DSS. In their endeavour, engineers have used a
range of scientific fields in physics, mathematics,
statistics, artificial intelligence and machine
learning.
These tools have not found intensive applications
so far (Han et al., 2002) that can vigorously utilise
their efficiency. This is referred in some extent to
that business and IT solution providers are
essentially interested in extensibility and
automation, and they aim at obtain fast results via
combining simple analysis with human expertise
(Ganguly and Gupta, 2005).
6 DM-BASED DSS SOLUTION
The vision of the Strategic and Performance
Development General Directorate of AD Police is to
transform the UAE into one of the safest and most
secure countries in the world. In this, they aim to
attain a more effective police force which responds
to the needs of society with the highest level of
integrity and training. For this aim, the organisation
is investing in new technologies to assist leaders in
decision-making and training, including Decision
Support Systems.
The abovementioned aim of the organisation to
make the country safer requires making decisions on
resource allocation. The available resources are
divided into two categories: policemen and financial
resources. These resources are to be allocated among
assets, such as policing roads and streets,
establishing new services, opening new departments
and sections, and building new police centres. The
decisions taken are concerned with the optimal
allocation of the available resources among the
assets. The DSS considers many variables including
but limited to number of crimes per area, crime
evolution, crime types, population per area,
population growth rate per area, national population
growth rate, available budgets, number of staff
members, balanced development, and general
guidance of the Directorate. These variables stem
from a multitude of data supplied to the system from
the different departments of the organisation (Figure
1).
Figure 1: Resource allocation among assets at AD Police
based on the information provided by the DSS.
Optimising the decision making process requires
targeting different stages of the process, including
the quality of the supplied data. This paper suggests
a framework to improve the DSS implemented in the
AD Police organisation in order to optimise the DSS
and hence the decisions made in the organisation.
Database management systems have advanced at
a faster rate than the techniques used for extracting
and utilising information to be used in making
decisions (Power, 2008), in the sense of using trends
of past endeavours to anticipate the future tendency
(Lv and Li, 2009). Obtaining, storing and managing
information in larger organisations are now ordinary
operations of business companies and usually
performed automatically by electronic data
repositories (Saxena and Rajpoot, 2009).
One of the efficient techniques used for this aim
(i.e. extracting useful information from large data) is
data mining. Some of the organisation’s information
may be in a textual format described in natural
languages; even systematically collected information
does not necessarily has a structure like the one used
for data tables and structured relational databases.
This type of information, found mainly in the form
of electronic documents and emails, particularly in
organisations with limited affordability (technical
and financial), cannot be used with traditional data
mining tools, and thus minimises the potential of
such organisations. The data miner thus has to
prepare the data for mining. This operation, called
pre-processing, is complex and may take several
months to complete, depending on the size of the
IntegrationofDecisionSupportSystemsandDataMiningforImprovedDecisionMaking
487
project, and require a significant part of the
organisational resources, and usually performed only
by large enterprises (Calderon et al, 2003). Artificial
Intelligence systems such as Neural Networks and
Fuzzy Inference algorithms might be appropriate for
finding the solution to this problem.
For police operations, however, the process
needs to be reduced in order to accommodate to the
particularity of AD Police in terms of size and
current system performance. For AD Police, Data
Mining Techniques (DMT) and DSS techniques can
be used to design and implement custom
applications that help in pattern recognition,
designing of predictive models, human-error
reduction, civil society’s protection, infrastructures
maintenance, quality enhancement and economic
and environmental sustainability (Ripley, 2008).
What we propose to enhance the currently used
DSS as opposed to the currently used DSS in the
organisation is improved implementation of DMT
based on extensive utilisation of DMT to attain
optimal results gained from advanced multi-
disciplinary combinations of these techniques. The
proposed system will require integration of a set of
techniques to be able to deliver the desired
outcomes. Those techniques consist of the database
system, the business logic and a user interface
model. The work will contribute to the development
and enhancement process through defining a
development approach and setting up all the needed
variables for establishing the solution. The proposed
DSS system will benefit from data mining processes
prior to having to deal with huge amounts of data
acquired from the different departments involved.
The techniques involved will be selectively
deployed on the different sets of data acquired
(Figure 2).
Figure 2: Components of the suggested DM-DSS system
for AD Police which are performed prior to the analysis
process stage by the DSS.
Several components will make up the data
mining system. The collected data will selectively
undergo several processes prior to be sent to the
DSS. These processes are data preparation,
classification, cluster analysis and association
analysis. Once done, the data enter into anomaly
detection process where anomalies in the processed
data will be detected and flagged up for further
investigation. These components are particularly
relevant to the AD Police DSS system given the data
formats and the other constraints such as time and
affordability limitations (Figure 3).
Figure 3: A diagram of the final DM-DSS depicting the
data mining processes that take place prior to the DSS
implementation stage. The DSS acquires relevant data to
support decisions on resource allocation.
7 CONCLUSIONS
This paper looked at the integration of DM-DSS
(Data Mining – Decision Support System) in the
literature in order to propose a solution for Abu
Dhabi Police. It was shown that the integration of
DM and DSS based solutions is not widely
implemented, although several researchers have
suggested the robustness of such integration. Given
the particular case of Abu Dhabi Police, the
currently used DSS can highly benefit from a DM-
DSS-based solution to improve the limited
performance and unsatisfactory results attained. The
suggested system uses DM as a pre-process to the
DSS implementation, all in an integrated DM-DSS.
The paper provided a prototype of DM-DSS
integrated system as part of the framework for AD
Police for in their data quality project. Given the
sparse data formats and the limited time and
affordability of the organisation, the suggested
prototype is tailored to the particularity of AD in
terms of size and DM techniques chosen. For the
implementation part of the prototype, it is necessary
to acknowledge that ‘Western’ processes usually
cannot be transferred directly. It is also necessary to
acknowledge that decision making process in the
West cannot be transferred directly to the case of
Abu Dhabi Police given the special aspects of Arab
culture. Therefore, cultural differences are
significantly influential in any system
implementation involving humans.
ICEIS2013-15thInternationalConferenceonEnterpriseInformationSystems
488
REFERENCES
Abu-Naser, S., Al-Masri, A., Sultan, Y. A., & Zaqout, I.
(2011). A Prototype Decision Support System for
Optimizing the Effectiveness of E-learning in
Educational Institutions. International Journal of Data
Mining & Knowledge Management Process, 1(4).
Calderon, T. G., Cheh, J. J., & Kim, I. (2003). How large
corporations use data mining to create value.
Management Accounting Quarterly, 4(2), 1-11.
El Seddawy, A. B., Khedr, A., & Sultan, T. Adapted
Framework for Data Mining Technique to Improve
Decision Support System in an Uncertain Situation.
Ganguly, A. R., & Gupta, A. (2005). Data mining
technologies and decision support systems for
business and scientific applications. Encyclopedia of
Data Warehousing and Mining, Blackwell Publishing.
Han, J., & Kamber, M. (2006). Data mining: concepts and
techniques. Morgan Kaufmann.
Han, J., Altman, R. B., Kumar, V., Mannila, H., &
Pregibon, D. (2002). Emerging scientific applications
in data mining. Communications of the ACM, 45(8),
54-58.
Hand, D. J., Mannila, H., & Smyth, P. (2001). Principles
of data mining. MIT press.
Hardin, J. M., & Chhieng, D. C. (2007). Data Mining and
Clinical Decision Support Systems. Clinical Decision
Support Systems, 44-63.
Hofstede, G. (2001). Culture's consequences: Comparing
values, behaviors, institutions and organizations
across nations. Sage Publications, Incorporated.
Kumar, D. S., Sathyadevi, G., & Sivanesh, S. Decision
Support System for Medical Diagnosis Using Data
Mining. International Journal of Computer Science, 8.
Liu, S., Duffy, A. H., Whitfield, R. I., & Boyle, I. M.
(2010). Integration of decision support systems to
improve decision support performance. Knowledge
and Information Systems, 22(3), 261-286.
Lv, J., & Li, Z. (2009, August). Research on architecture
of decision support system. In Computer Science and
Information Technology, 2009. ICCSIT 2009. 2nd
IEEE International Conference on (pp. 199-202).
IEEE.
Markov, Z., & Larose, D. T. (2007). Data mining the
Web: uncovering patterns in Web content, structure,
and usage. Wiley-Interscience.
McCue, C., & Parker, A. (2003). Connecting the Dots:
Data Mining and Predictive Analytics in Law
Enforcement and Intelligence Analysis. Police Chief,
70(10), 115-124.
Mladenić, D. (2001). EU project: Data mining and
decision support for business competitiveness: a
European virtual enterprise (Sol-Eu-Net). OES-SEO
2001: Open enterprise solutions: Systems, experiences
and organizations, 172-173.
Mladeni
ć, D. (2003). Data mining and decision support:
Integration and collaboration. Springer.
Mohemad, R., Hamdan, A. R., Othman, Z. A., & Noor, N.
M. M. (2010). Decision support systems (dss) in
construction tendering processes. arXiv preprint
arXiv:1004.3260.
Oatley, G., Ewart, B., & Zeleznikow, J. (2006). Decision
support systems for police: Lessons from the
application of data mining techniques to “soft”
forensic evidence. Artificial Intelligence and Law,
14(1), 35-100.
Padhy, N., Mishra, P., & Panigrahi, R. (2012). The Survey
of Data Mining Applications And Feature Scope.
International Journal of Computer Science.
Power, D. J. (2008). Understanding data-driven decision
support systems. Information Systems Management,
25(2), 149-154.
Ripley, B. D. (2008). Pattern recognition and neural
networks. Cambridge university press.
Saxena, K., & Rajpoot, D. S. (2009). A Way to
Understand Various Patterns of Data Mining
Techniques for Selected Domains. arXiv preprint
arXiv:0911.0781.
Segall, R. S., & Zhang, Q. (2006, March). Applications of
Neural Network and Genetic Algorithm Data Mining
Techniques in Bioinformatics Knowledge Discovery–
A Preliminary Study. In Proceedings of the Thirty-
seventh Annual Conference of the Southwest Decision
Sciences Institute (Vol. 37, No. 1).
Srinivasan, S., Singh, J., & Kumar, V. Multi-agent based
decision Support System using Data Mining and Case
Based Reasoning. International Journal of Computer
Science, 8.
Wang, C. Y., Tseng, S. S., & Hong, T. P. (2006). Flexible
online association rule mining based on
multidimensional pattern relations. Information
Sciences, 176(12), 1752-1780.
IntegrationofDecisionSupportSystemsandDataMiningforImprovedDecisionMaking
489
