<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Archiving and Interchange DTD v1.0 20120330//EN" "JATS-archivearticle1.dtd">
<article xmlns:xlink="http://www.w3.org/1999/xlink">
  <front>
    <journal-meta />
    <article-meta>
      <title-group>
        <article-title>Knowledge Management on a Strategy for Requirements Engineering (KMoS-RE) and Case-Based Reasoning (CBR) Applied in Cooling Unit Design</article-title>
      </title-group>
      <contrib-group>
        <contrib contrib-type="author">
          <string-name>Yadira Kiquey Ortiz Chow</string-name>
          <email>yortiz@uacj.mx</email>
          <xref ref-type="aff" rid="aff1">1</xref>
        </contrib>
        <contrib contrib-type="author">
          <string-name>Pedro Bañuelos Aguilar</string-name>
          <email>pedro.banuelos@uacj.mx</email>
          <xref ref-type="aff" rid="aff2">2</xref>
        </contrib>
        <contrib contrib-type="author">
          <string-name>Jorge Rodas-Osollo</string-name>
          <email>jorge.rodas@uacj.mx</email>
          <xref ref-type="aff" rid="aff0">0</xref>
        </contrib>
        <contrib contrib-type="editor">
          <string-name>Case-Based Reasoning; CBR; KMoS-RE</string-name>
        </contrib>
        <aff id="aff0">
          <label>0</label>
          <institution>Laboratorio Nacional en Tecnologías</institution>
          ,
          <addr-line>de la Información, sede UACJ</addr-line>
        </aff>
        <aff id="aff1">
          <label>1</label>
          <institution>Maestría en Cómputo</institution>
          ,
          <addr-line>Aplicado.DEyC, IIT.UACJ</addr-line>
        </aff>
        <aff id="aff2">
          <label>2</label>
          <institution>Maestría en Cómputo</institution>
          ,
          <addr-line>Aplicado.DEyC, IIT.UACJ</addr-line>
        </aff>
      </contrib-group>
      <abstract>
        <p>Every day it can be observed that there are situations that are resolved based on each person experience, which is obtained by solving situations that happened before whether the solutions were correct o incorrect. Therefore, the CBR is widely used for solving problems that are distinguished by the need to use the experience gained previously. On the other hand, there are situations where you need to have well defined what you want to do or what it is asked to be made, for that, the KMoS-RE strategy helps to define the requirements. In the productive sector, having specific requirements and the experts experience is very useful for the projects to be designed, developed and delivered as they were requested.</p>
      </abstract>
    </article-meta>
  </front>
  <body>
    <sec id="sec-1">
      <title>1. INTRODUCTION</title>
      <p>Nowadays to provide a good quality service or produce a product,
it is necessary to know the needs and characteristics of the problem
that is intended to be solved; those features only the person or
people who know the needs can express them. Although this does
not assure that the needs are the ones to be solved; since often the
final user has the tacit knowledge and is difficult to explain what
he needs. The KMoS-RE strategy is focused on making the
transformation of tacit knowledge to explicit knowledge through a
series of activities that helps the requirements engineer obtain the
expert knowledge or the user who has the need. In addition, the
Case-Based Reasoning is a tool that simulates the way humans
solve problems, because usually a human being builds on its
experience how to solve new problems that have similarities with
the previous problems.</p>
      <p>In this article KMoS-RE methodology and CBR are combined to
give a solution to the needs of the company Flutec, which is related
to the design and manufacture of cooling units; resulting in a
customer satisfaction when results a shown.</p>
    </sec>
    <sec id="sec-2">
      <title>2. BACKGROUND</title>
      <p>
        The Cased-Based Reasoning (CBR) focuses in the problem
resolution, having a baseline the previous solved problems and the
solution to these problems. In this type of resolution it is very
similar as the way humans resolved the day to day problems,
whether is it conscious or unconscious humans always remember
the similar situations before making a decision or giving a solution
to the problem. [
        <xref ref-type="bibr" rid="ref1">1</xref>
        ]
Bear in mind, it can be said that the case-based reasoning (RBC) it
is a problem solver oriented paradigm, which needs a previous
knowledge of the domain that is known as experience and relying
with the previous solutions of the problems, doing the necessary
comparisons so that gets to the optimal solution for the problem.
The CBR has a cycle that follows in order to give a solution to a
new problem in the domain we are dealing, it is divided in four
parts, which are: retrieve, reuse, revise and retain. In order to begin
the CBR cycle, it is needed a case base, which means that is
essentially to have represented the previous problems set. To do
this, it is said that the problem is represented in a case. Generally,
in the figure 1 it can be observed the RBC cycle.
With the use of Case Base Reasoning it is pretend to model the
knowledge that the experts had in the domain that is going to be
applied in the CBR.To do so, it is necessary to implement together
with the CBR another tool that help us obtain the knowledge of the
experts, on certain occasions the language used by the experts it is
not understood at all by the requirements engineers and this can
lead into suppositions or assumptions things that don’t belong in
the domain. Actually, there are already tested tools for obtained an
experts knowledge an one of them is the Knowledge Management
on a Strategy for Requirements Engineering (KMoS-RE) [3][
        <xref ref-type="bibr" rid="ref3">4</xref>
        ] is
it a strategy design for create, structured and obtain knowledge in
domains ill-structured, than can be in incorporated to a problems
solution and avoid incomplete and ambiguous requirements. In the
figure 2 shows the general way that KMoS-RE consists. Later it is
detailed each of the artefacts that conforms a part in the strategy.
KMoS-RE strategy consist of:
 Record of beliefs: In this point, annotations of
assumptions are taken by the requirement engineers that
they have about the domain in which they are working,
with the purpose that if a new member is added to the
work team, the new member does not formulate the





same assumptions and his integration is easier to the
domain.
      </p>
      <p>Knowledge of Domain on an Extended Lexical
(KDEL): To generate this artifact, interviews are made
to the experts, then analyze the info, which means the
requirement engineers listens it, generate notes and then
ask questions to the experts in the domain. With that
being said, the requirement engineers are capable to
obtain concepts, definitions and relations used in the
domiain so that they can elaborate the conceptual
model.
Conceptual Model: Is generated based on the KDEL,
which is showed to the domain expert and he validates
the information contained in the model. In case there is
some inconformity, it is notified to the requirement
engineer and he does the modification.</p>
      <p>Current use case model: Contains the functions of the
current system, which means, the way the things are done
actually. To get an idea on how it is conformed and the
used of each terms that are in the KDEL.</p>
      <p>Future use case model: This is the actual system
enhanced, these adds o eliminates functions to the current
system.</p>
      <p>Specification document: Contains all the required
information to develop the system that the expert is
requesting
There is a domain ill-structured belonging to Flutec Design + Build
Heating Ventilation and Air Conditioning (HVAC), Flutec is a
company that offers their services in an international level, their
mainly activity is the design and custom made production of
chillers, Chil-Paks, vapor and air compressed modules. All of the
mentioned is offered to a company level.</p>
      <p>Off all the product types that Flutec elaborates, the boarded projects
are the ones relative to the cooling units (chil-paks). In used terms
of the domain, a cooling unit is a central plant integrated, packed in
a plug and play unit. To develop each cooling unit, the engineering
department must meet a set of characteristics or features that are
considered unrepeatable and unique, because it relies heavily in the
use they give to the cooling unit, for this reason it is ideal that these
characteristics or features are granted by the person responsible by
the client, in the majority of cases it is a mechanical engineer the
person responsible. Taking all this into consideration, to develop a
proposal must be done independently, considering that the amount
of information is huge; in the same way it must be contemplate the
knowledge gain by the experience of the team work working in the
project.</p>
      <p>Actually Flutec uses a document that is called DNA, this is their
support tool for obtaining a better definition of the chil-pak that a
client requires and with this ensure that the propose given to the
client meets with the expectations. One thing must be remembering
in order to develop any type project, if the specification or the
characteristics are not well defined, ambiguous or it is not well
understood what the client is needing, the final product will be
something different of what the client is requesting. Once the DNA
is ready it is considered an empiric requirement document and it is
taken like a base for the design and construction for the chil-pak.
Flutec currently uses a document that is called DNA, this is their
tool support for better definition of the cooling unit that the
customer needs and thereby to ensure that the proposal will be
given to the customer meets the same expectations. It is worth
remembering that in order to develop any project, whether the
specification or the features are not well defined, no ambiguities or
misunderstanding of what the customer is asking, the final product
will be something different than what the client is requesting. Once
the DNA is ready it is considered as an empirical document of
requirements and is taken as a basis for the design and construction
of cooling units.</p>
      <p>To Flutec is of utmost importance to define in detail the cooling
unit that the client has in mind, as Flutec staff invest time and
resources for the generation of proposals that are sent to the client
for assessment, this is done in order to obtain the development of
the cooling unit. Therefore, the problem that arises in Flutec is
when they generate the proposal, because the DNA is currently
filled by several people within the company, and sometimes require
the intervention of the engineering department. By requiring a long
time to develop and submit a proposal, It is in this step where it
causes a great disadvantage to Flutec against their competitors
market. In addition to elaboration time, other factors affect the
delivery of the proposal as it is, the knowledge is distributed among
experts of the company, this factor affected when one of the experts
is absent and there is no one who can replace it at that time. Another
factor is teamwork, as there is a very strong dependence on the
engineering department and this depends on many characteristics
of the cooling unit.</p>
      <p>Performing a domain analysis and the information handled in
Flutec, it was founded that Flutec made custom made water cooling
units (chil-pak), which means that a chil-pak is generated based on
the specific needs for each client. Therefore, Flutec staff and sellers
use the document called DNA to collect customer information and
make this a proposal to the chil-pak containing the specification
given by the client. The higher definition it has of what the
customer requires, better product will be delivered and the lower
the probability that the product delivered has any errors, with this
Flutec meets customer needs.</p>
      <p>At present, although the DNA was conceived as a document to
assist in the definition of the cooling unit, this document is of little
use because of its complexity filling. In addition, people who
interact with DNA at the beginning of the project definition
unknow the meaning and use of information that is required in the
DNA, leaving the document incomplete and forcing the interaction
with others in order to finish the DNA filling. This is reflected in
time consumption and results in missing opportunities that Flutec
may keep with the development of the cooling unit.</p>
      <p>Studying the information provided by the company through
KMoSRE strategy and knowing the characteristics for using the case
based reasoning, it was chosen to represent knowledge using the
pair attribute - value, and use the CBR to give a solution to the need
of the company.</p>
    </sec>
    <sec id="sec-3">
      <title>3. DEVELOPMENT PROCESS</title>
      <p>It started with an interview to become familiar with the domain in
which it will work. After that, a model of the domain was done and
it was validated by engineering coordination, since they are the
ones who generate proposals. Then it was performed a series of
interviews with the general director and the engineering
coordination in order to complement the domain model that had
been done.</p>
      <p>It was observed that for handling a DNA there are different
approaches over its importance. On one hand, it is believed that if
you do not have a DNA it could not get a good definition of the
cooling unit desired by the customer and that the proposal
submitted may contain ambiguities that affected at the time the
proposal is authorized; seeing the other point of view that they have
of the DNA within Flutec , it is consider very tedious filling and
also is not much help for the project, as they only take a few data
from the DNA for the preparation of the proposal and the
construction of the cooling unit. Therefore, as they cannot see a use
of the DNA, they left it out or simply filled when the proposal was
generated.</p>
      <p>The way to apply the KMoS-RE strategy consisted in recording the
interviews made with the general director and engineering
coordination of Flutec were made, subsequently the interviews
were transcribed and then it began to work with the domain
modeling phase of the corresponding in the strategy KMoS-RE. In
this first phase, the extended lexicon of language (KDEL) was
made, in other words, terms, symbols, concepts and relationships
of the domain were obtained. Subsequently, the conceptual model
was elaborated and validated by the engineering coordination was
conducted. Realizing the first phase of the strategy, helped us to
know the domain and to better understand the terms used by the
domain experts, because when the project started the cooling units
was a domain completely unknown.</p>
      <p>The information gained in the interviews is:
• The explanation of Flutec's business activity.
• The DNA importance for the definition of a water
cooling unit.
• The use of the DNA for the proposal generation.
• The working process in FLutec for developing a
chilpak.
• Flutec's organization.
• The meaning of each phase used in the DNA.
• The basic information for generate a proposal or a
chilpak.</p>
      <p>The information contributed by Flutec is:
•
•
•
•
•
•
•
•
•
•
•</p>
      <p>Access to the DNA's registered in their actual system.
Filling DNA information belonging to a project.</p>
      <p>Visualization of a PIN ID and MH100 plan.</p>
      <p>Organization of the physical information of a project.</p>
      <p>Phases in which the DNA's is organized.</p>
      <p>When the information it was already had, an analysis was done, in
which case it realizes that a DNA besides being used for defining
the chil-pak, also is used to track the project, which means, is used
to know the stage that the project it is. The aim is to know what
projects are in current guaranteed.</p>
      <p>Taking the basic information for generating a proposal, it was
chosen to use the CBR to give a solution to the need of the
company. That led to implement the RBC in the company. Since
basically the Flutec Company requires decreasing the time that
takes to generate a proposal and this ranges from the time the
specifications are sent, through filling DNA and arriving in the first
instance to the proposal that is sent to the client approval. That led
to undertake the development of case-based reasoning to solve the
need that the company has Flutec, the way it was working was:
Having the domain model that was obtained from the
KMoSRE strategy implementation, the most important attributes for
the engineering coordination were obtained, in example, the
most significant data were taken from the approach of
engineering coordination that they were useful to them in
order to compare or select previous projects.</p>
      <p>After that it was continued by assigning them the importance
to the attributes, with the purpose to detect the attribute that
has the highest priority for developing the proposal of a
chilpak.</p>
      <p>Because the Case-Based Reasoning requires a case base to
begin with its cycle, it was necessary for the case
representation with the attribute-value combination; so that
these cases were introduced into the case base.</p>
      <p>Starting the Case-Based Reasoning cycle, for the recovery
case phase, the formula of nearest neighbor similarity was
implemented using the priority that was assigned to each
attribute marked in the representation cases.</p>
      <p>By getting the most similar cases to the new problem, experts
consulted to inform if the result was correct.</p>
      <p>Lately, the adaptation of the case was made and the domain
experts decided if the solution given by the CBR is included
or not in the case base.</p>
      <p>Presenting the implementation of the CBR to the coordination
engineering, it was realized that this implementation met the
expectations that the company had regarding the project of Case
Based Reasoning, since it was easier for them to detect if previously
has already been made a project similar to the requirements of the
new project requirements; with the purpose to obtain the plans and
specifications of the equipment used in the solution of the
mentioned project, thereby develop in less time the proposal so that
it can be sent to the client.</p>
    </sec>
    <sec id="sec-4">
      <title>4. RESULTS</title>
      <p>At the moment of implementing the KMoS-RE strategy to structure
their knowledge experts in Flutec, it was obtained the extended
lexicon of language (KDEL), the conceptual model and the domain
model. Latter on it has two views, one from the coordination
engineering (Figure 3) and the other from the general direction
(Figure 4).</p>
      <p>An important part of implementing the strategy, is that people
outside the domain acquired a basic knowledge of the domain more
quickly, helping that the communication with the experts were
smoother. Similarly, it can focus on the problem area, which means,
it helps requirements engineers to delimit their questions or to
address specific concerns of the domain, thereby transform as much
tacit knowledge of the expert into explicit knowledge.
After carrying out activities from KMOS-RE strategy, the
CaseBased Reasoning was implemented as a solution to the need of
Flutec; The Service of Quality (SERVQUAL) questionnaire was
applied to the persons involved in the project to meet the
expectation and perception of the company to the service that was
given. Within this questionnaire the questions are classified into 5
categories: tangible, reliability, response, warranty and empathy.
The minimum score that can be obtained in response to each
question is a number 1, which means that the respondent
completely disagreed with was is stated in the question; the highest
score you can get is a number 7 which represents the respondent
completely agrees with what is stated in the question.
which means, the service surpassed the expectations of the
company. This can be seen in table 1.</p>
      <p>Table 1 SERVQUAL Questionnaire application results.
On the other hand, the Software Usability Measurement Inventory</p>
      <sec id="sec-4-1">
        <title>Average</title>
        <p>expectation</p>
      </sec>
      <sec id="sec-4-2">
        <title>Average</title>
        <p>perception</p>
      </sec>
      <sec id="sec-4-3">
        <title>Category</title>
      </sec>
      <sec id="sec-4-4">
        <title>Tangible</title>
      </sec>
      <sec id="sec-4-5">
        <title>Reliability</title>
      </sec>
      <sec id="sec-4-6">
        <title>Response</title>
      </sec>
      <sec id="sec-4-7">
        <title>Warranty</title>
      </sec>
      <sec id="sec-4-8">
        <title>Empathy</title>
      </sec>
      <sec id="sec-4-9">
        <title>Category</title>
        <sec id="sec-4-9-1">
          <title>Accuracy</title>
        </sec>
        <sec id="sec-4-9-2">
          <title>Usability</title>
        </sec>
        <sec id="sec-4-9-3">
          <title>Visibility and</title>
          <p>documentation</p>
        </sec>
        <sec id="sec-4-9-4">
          <title>Understandability</title>
        </sec>
        <sec id="sec-4-9-5">
          <title>Reliability Efficiency</title>
          <p>1.00
6.20
6.75
7.00
7.00
4.33
6.20
6.50
6.75
6.40
(SUMI) questionnaire was also applied in order to measure the
expectation and perception of experts Flutec as the prototype was
delivered. For this questionnaire the questions were classified into
6 categories: accuracy, usability, documentation and visibility,
understandability, reliability and efficiency. In the same way that
the SERVQUAL questionnaire application, the minimum score to
answer a question is the number 1 and the maximum score is
number 7 , where the number 1 represents that the respondent is
completely disagree and the number 7 represents the respondent is
completely agree.</p>
          <p>After performing the application of the questionnaire and analyze
the results it is concluded that the prototype of CBR fulfilled with
88.65 % of customer expectations; this data is obtained by
considering the data shown in Table 2.</p>
        </sec>
      </sec>
    </sec>
    <sec id="sec-5">
      <title>5. CONCLUSION</title>
      <p>After carry out the implementation of the KMOS -RE strategy and
the Case-Based Reasoning (CBR), it was concluded that not only
in software development the requirements are used and that these
requirements need to be well defined to avoid ambiguity and errors
in the construction of software or any final product. All off this in
order that the final user is satisfied with the delivered product. In
the vast majority of projects the definition of requirements it is not
given the proper importance and therefore, the last thing you want
is to spend time detailing the requirements because the customer
needs is the product as soon as possible.</p>
      <p>Unfortunately, when the physical development of the product gets
more importance than the requirements, many things will be
assumed and in the future these suppositions can cause re-work,
spend more time and that the final user is not satisfied with the
delivered product.</p>
      <p>For a company such as Flutec it represents a great disadvantage the
fact that they depend on knowledge of its experts, and that they do
not know at what point these experts will stop working for the
company and carried them all the tacit knowledge acquired.
Therefore, implementing strategies such as KMoS-RE for tacit
knowledge of experts and the paradigm for Case-Based Reasoning
helps this disadvantage diminishes greatly.</p>
      <p>Speaking of Case-Based Reasoning for solving the need for the
Flutec’s company, it was observed that they perform the same CBR
cycle unconsciously, relying on memory, knowledge and
experience of its experts in manufacturing cooling units. Because
the results obtained by the CBR corresponded with the results
obtained by performing experts latter cycle empirically.</p>
    </sec>
    <sec id="sec-6">
      <title>6. FUTURE WORK</title>
      <p>Since two points of view are shown to generate a proposal and a
water cooling unit, it would be interesting to carry out an
experiment in which both the General Director and the engineer
coordination work on a project each in their own way, to determine
if they come to the same solution and evaluate which option was
the best, in other words, evaluate execution time and evidence
generated in each of the options.</p>
      <p>Additionally, it was observed that there is a strong dependence on
the engineering coordinator which leads us to recommend that a
transitional simulation must be done, so that the person currently in
charge of projects trains new member that stays in that position.
When we have important members in a company, it must have
contemplated the benefits and disadvantages that brings, because
that depends on the economic stability of the company.
If another use was given to the DNA and the information contain
was reformulated, they could be forced to use it. Because it is
always important to have a good requirements specification before
develop or construct any product. Most of the time requires
considerable time to achieve a good specification, but this will
bring benefits as reduction of construction time for the company
this can translates into money as it can acquire more projects. It can
be seen it from the customer's perspective, which will be satisfied
with the product delivered and it will meet the needs expressed.</p>
      <sec id="sec-6-1">
        <title>Other uses that the DNA could have are: • • •</title>
        <p>A checklist, which means after the complete construction of
the cooling unit it could be used to verify that all the
specifications given by the client are in the cooling unit.
Use the DNA information for statistical production.</p>
        <p>A Tool that helps to track the project progress, which means,
if a proposal or project is consulted the information contained
was up to date.</p>
        <p>In addition, if all the projects information is available this would be
used as a base for developing future proposals and projects. With
this, the process of generating proposal does not depend on the
knowledge of one person.</p>
      </sec>
    </sec>
  </body>
  <back>
    <ref-list>
      <ref id="ref1">
        <mixed-citation>
          [1]
          <string-name>
            <given-names>. A.</given-names>
            <surname>Aamodt</surname>
          </string-name>
          ,
          <string-name>
            <given-names>E.</given-names>
            <surname>Plaza</surname>
          </string-name>
          (
          <year>1994</year>
          ) “
          <article-title>Case-Based Reasoning: Foundational Issues, Methodological Variations, and System Approaches”</article-title>
          .
          <source>AI Comminications</source>
          . IOS Press, Vol.
          <volume>7</volume>
          :
          <issue>1</issue>
          , pp.
          <fpage>39</fpage>
          -
          <lpage>59</lpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref2">
        <mixed-citation>
          [2]. A.
          <string-name>
            <surname>Fornells</surname>
          </string-name>
          (
          <year>2009</year>
          )
          <article-title>“A Unified Framework for the Development of Case-Based Reasoning Systems”</article-title>
          . Recuperado el 23 de Marzo de 2016, de http://lasallerd.salleurl.edu/A_Unified_
          <article-title>Framework_for_the_ Develop-ment_</article-title>
          <source>of_Case-Based_Reasoning_Systems%2C [3]. Olmos-Sánchez Karla… [et all]</source>
          . (
          <year>2015</year>
          ) “
          <article-title>Requirements engineering based on knowledge: a comparative case study of the KMoS-RE strategy and the DMS process”</article-title>
          . Revista Facultad de Ingeniería, Universidad de Antioquia,
          <source>No. 77</source>
          , pp.
          <fpage>88</fpage>
          -
          <lpage>94</lpage>
          ,
          <year>2015</year>
          .
        </mixed-citation>
      </ref>
      <ref id="ref3">
        <mixed-citation>
          [4]
          <string-name>
            <given-names>. K.</given-names>
            <surname>Olmos</surname>
          </string-name>
          and
          <string-name>
            <given-names>J.</given-names>
            <surname>Rodas.</surname>
          </string-name>
          KMoS-RE:
          <article-title>knowledge management on a strategy to requirements engineering</article-title>
          .
          <source>Requirement Engineering Journal</source>
          .
          <year>2014</year>
          . ISSN 0947-
          <issue>3602</issue>
          , Volume =
          <volume>19</volume>
          , Number = 4, Springer-Verlag New York, Inc.
        </mixed-citation>
      </ref>
      <ref id="ref4">
        <mixed-citation>
          [5]
          <string-name>
            <given-names>. M.</given-names>
            <surname>Behbahani</surname>
          </string-name>
          ,
          <string-name>
            <given-names>A.</given-names>
            <surname>Saghaee</surname>
          </string-name>
          ,
          <string-name>
            <given-names>R.</given-names>
            <surname>Noorossana</surname>
          </string-name>
          (
          <year>2012</year>
          )
          <article-title>“A casebased reasoning system development for statistical process control: Case representation and retrieval”</article-title>
          . Computer &amp; industrial Engineering, Elsevier. Vol.
          <volume>63</volume>
          pp.
          <fpage>1107</fpage>
          -
          <lpage>1117</lpage>
          [6]
          <string-name>
            <given-names>. H.</given-names>
            <surname>Li</surname>
          </string-name>
          ,
          <string-name>
            <given-names>J.</given-names>
            <surname>Sun</surname>
          </string-name>
          (
          <year>2009</year>
          ) “
          <article-title>Predicting business failure using multiple case.based reasoning combined with support vector machine”</article-title>
          .
        </mixed-citation>
      </ref>
      <ref id="ref5">
        <mixed-citation>
          <source>ELSEVIER Expert Systems with Applications</source>
          Vol.
          <volume>36</volume>
          pp.
          <fpage>10085</fpage>
          -
          <lpage>10096</lpage>
        </mixed-citation>
      </ref>
      <ref id="ref6">
        <mixed-citation>
          [7]. S. Ji,
          <string-name>
            <given-names>M.</given-names>
            <surname>Park</surname>
          </string-name>
          ,
          <string-name>
            <given-names>H.</given-names>
            <surname>Lee</surname>
          </string-name>
          (
          <year>2012</year>
          )
          <article-title>“Case Adaptation Method of Case- based Reasoning for Construction Cost Estimation in Korea”</article-title>
          .
          <source>Journal of Construction Engineering and Management ©ASCE</source>
          Vol.
          <volume>138</volume>
          No. 1 pp.
          <fpage>43</fpage>
          -
          <lpage>52</lpage>
        </mixed-citation>
      </ref>
    </ref-list>
  </back>
</article>