Docstoc

A Decision Tree Based Model to Identify the Career Focus of Computer Stream Students in ITES Industry

Document Sample
A Decision Tree Based Model to Identify the Career Focus of Computer Stream Students in ITES Industry Powered By Docstoc
					                                                        (IJCSIS) International Journal of Computer Science and Information Security,
                                                        Vol. 9, No. 9, September 2011




   A Decision Tree Based Model to Identify the
   Career Focus of Computer Stream Students in
                  ITES Industry

                                        T.Hemalatha#1, Dr.Ananthi Sheshasaayee #2

 #1. T.Hemalatha, Research Scholar, R&D centre, Bharathiar University,Combatore, Asst.Prof, M.C.A.Dept,VELS
                                          University,Chennai,India
 #2. Dr.Ananthi Sheshasaayee- Associate Professor and Head, Dept of Computer Science, Quaid-E-Millath Govt.
                              College for women (Autonomous), Chennai –600 002,India

                                               1
                                                   winhema18@yahoo.co.in
                                                    2
                                                    ananthiseshu@gmail.com

Abstract - This paper focuses on the various career                    is the decision tree ,which is utilized in this paper.
opportunities that are available for the computer                      Decision tree learning is one of the most successful
stream students in the field of ITES industry .This                    learning algorithms, for its various attractive features.
paper analyses the various attributes of the Skill Set of              Simplicity, comprehensibility, parameter less, and
computer stream students, from which a decision tree
                                                                       being able to handle mixed type data. In decision tree
can be generated to help them to improve the
confidence level of students in selecting a career in ITES             learning, a decision tree is induced from a set of
industry. For the past few years it has become a passion               labeled training instances represented by a tuple of
for students to choose computer science as their main                  attribute values and a class label. Because of the vast
stream for their studies. During the final semester of                 search space, decision tree learning is typically a
their graduation they struggle a lot to choose a career                greedy,top-down and recursive process starting with
based on the skill set they posses which is of due                     the entire training data and an empty tree. An
importance. With the use of Decision tree this paper                   attribute that best partitioned into disjoint subsets
provides a guideline to take decision to choose career in              satisfying the values of the splitting attribute, for each
ITES Industry.
                                                                       subset, the algorithm proceeds recursively until all
Keywords - skill set, career, computer stream, ITES,                   instances in a subset belong to the same class [1].
decision tree, decision                                                          Decision trees are a rapid and effective
                                                                       method of classifying data set entries, and can offer
                 I. INTRODUCTION                                       good decision support capabilities. A decision tree is
                                                                       a tree in which each non-leaf node denotes a test on
          In this competitive world it is very difficult               an attribute of cases, each branch corresponds to an
to secure a job .Students who have chosen computer                     outcome of the test, and each leaf node denotes a
as their main stream can decide their career based on                  class prediction. The quality of a decision tree
the skill set they posses. They are not much aware                     depends on both its classification accuracy and its
about the skills required by the ITES industry.                        size.[2] Existing studies have identified several
Knowing the skills they possess, we can give them a                    advantages to the use of decision trees: no domain
decision to choose their career.                                       knowledge is needed for classification, they are able
          In our day today life, we come across                        to handle high dimensional data, they are intuitive
various decision making problems. Normally we                          and generally easy to comprehend, they are simple
solve these problems and make decisions out of the                     and fast, and they have good accuracy [3].
experience which may be incorrect seldom. The
computer technology helps us to provide an easy and
efficient way of decision making. One such approach




                                                                 91                                http://sites.google.com/site/ijcsis/
                                                                                                   ISSN 1947-5500
                                                 (IJCSIS) International Journal of Computer Science and Information Security,
                                                 Vol. 9, No. 9, September 2011




            II. RESEARCH METHODS                                Communication skill 2. Knowledge on productivity
                                                                software 3. Domain Knowledge 4. Soft Skill 5
     1. Data collection                                         Decision making 6. Analytical skills..
     Data for this study were collected from various
ITES workers like developers, web designers, System                        III. EXPERIMENTAL RESULTS
admin, Team lead, Project Manager, developer,
Testers, copy editor, reference setting, Hr, network                      Based on the answer given by the them, the
admin, etc. We contacted many workers from ITES                 important attribute is selected on the basis of highest
industry and a questionnaire was given to them The              percentage and then attribute was categorized by the
questionnaire is of closed type. Participation in this          requirement needed by the ITES industry. It has been
study was voluntary and people were assured that                identified that the major skills set which are required
their individual responses would be treated as                  for BPO companies are Knowledge on productivity
confidential.                                                   software, Communication skill, Domain Knowledge,
                                                                Soft Skill and for KPO companies, Knowledge on
                                                                productivity software, Communication skill, Domain
    2.   Data Set Description                                   Knowledge, Soft Skill, Decision making, Analytical
                                                                skills are required. We Assign Knowledge on
          Among the various attributes the following            productivity software as KPS, communication skill
attributes are considered as vital for career decision          as CS, Domain Knowledge as DK, Soft skill as SS,
in ITES Industry. The vital attributes are 1.                   Decision Making as DM and Analytical skill as AS.

Table I. Data Set for the ITES Industry
 SAMPLE      CS          KPS         DK          SS            DM           AS           BPOEF         KPOEF
 P1                  1           1           1            1            1             1            1              1
 P2                  1           1           1            1            1             0            1            0.8
 P3                  1           1           1            1            0             1            1            0.8
 P4                  1           1           1            1            0             0            1            0.7
 P5                  1           1           1            0            1             1          0.8            0.8
 P6                  1           1           1            0            1             0          0.8            0.7
 P7                  1           1           1            0            0             1          0.8            0.7
 P8                  1           1           1            0            0             0          0.8            0.5
 P9                  1           1           0            1            1             1          0.8            0.8
 P10                 1           1           0            1            1             0          0.8            0.7
 P11                 1           1           0            1            0             1          0.8            0.7
 P12                 1           1           0            1            0             0          0.8            0.5
 P13                 1           1           0            0            1             1          0.5            0.7
 P14                 1           1           0            0            1             0          0.5            0.5
 P15                 1           1           0            0            0             1          0.5            0.5
 P16                 1           1           0            0            0             0          0.5            0.3
 P17                 1           0           1            1            1             1          0.8            0.8
 P18                 1           0           1            1            1             0          0.8            0.7
 P19                 1           0           1            1            0             1          0.8            0.7
 P20                 1           0           1            1            0             0          0.8            0.5
 P21                 1           0           1            0            1             1          0.5            0.7
 P22                 1           0           1            0            1             0          0.5            0.5
 P23                 1           0           1            0            0             1          0.5            0.5
 P24                 1           0           1            0            0             0          0.5            0.3
 P25                 1           0           0            1            1             1          0.5            0.7
 P26                 1           0           0            1            1             0          0.5            0.5
 P27                 1           0           0            1            0             1          0.5            0.5




                                                          92                                http://sites.google.com/site/ijcsis/
                                                                                            ISSN 1947-5500
                                                  (IJCSIS) International Journal of Computer Science and Information Security,
                                                  Vol. 9, No. 9, September 2011




 P28                 1            0           0            1            0             0          0.5            0.3
 P29                 1            0           0            0            1             1          0.3            0.5
 P30                 1            0           0            0            1             0          0.3            0.3
 P31                 1            0           0            0            0             1          0.3            0.3
 P32                 1            0           0            0            0             0          0.3            0.2
 P33                 0            1           1            1            1             1          0.8            0.8
 P34                 0            1           1            1            1             0          0.8            0.7
 P35                 0            1           1            1            0             1          0.8            0.7
 P36                 0            1           1            1            0             0          0.8            0.5
 P37                 0            1           1            0            1             1          0.5            0.7
 P38                 0            1           1            0            1             0          0.5            0.5
 P39                 0            1           1            0            0             1          0.5            0.5
 P40                 0            1           1            0            0             0          0.5            0.3
 P41                 0            1           0            1            1             1          0.5            0.7
 P42                 0            1           0            1            1             0          0.5            0.5
 P43                 0            1           0            1            0             1          0.5            0.5
 P44                 0            1           0            1            0             0          0.5            0.3
 P45                 0            1           0            0            1             1          0.3            0.5
 P46                 0            1           0            0            1             0          0.3            0.3
 P47                 0            1           0            0            0             1          0.3            0.3
 P48                 0            1           0            0            0             0          0.3            0.2
 P49                 0            0           1            1            1             1          0.5            0.7
 P50                 0            0           1            1            1             0          0.5            0.5
 P51                 0            0           1            1            0             1          0.5            0.5
 P52                 0            0           1            1            0             0          0.5            0.3
 P53                 0            0           1            0            1             1          0.3            0.5
 P54                 0            0           1            0            1             0          0.3            0.3
 P55                 0            0           1            0            0             1          0.3            0.3
 P56                 0            0           1            0            0             0          0.3            0.2
 P57                 0            0           0            1            1             1          0.3            0.5
 P58                 0            0           0            1            1             0          0.3            0.3
 P59                 0            0           0            1            0             1          0.3            0.3
 P60                 0            0           0            1            0             0          0.3            0.2
 P61                 0            0           0            0            1             1          0.0            0.3
 P62                 0            0           0            0            1             0          0.0            0.2
 P63                 0            0           0            0            0             1          0.0            0.2
 P64                 0            0           0            0            0             0          0.0            0.0

In the Table I. strong skill set value is represented as         software, domain knowledge, soft skills .ie.SOSS1=
1 and the weak skill set value is represented as 0.The           ∑ [(cs) +(kps) +(dk)+(ss)
value ranges for BPO Key Factor and KPO Key                      SOSS2= Summation of the all six attributes they are
factor varies from 0.0 to 1.0.                                   communication skills, knowledge on productivity
BPO Key Factor {0.0 – 1.0}                                       software, domain knowledge, Soft Skill ,decision
KPO Key Factor {0.0 -1.0}                                        making, Analytical skills i.e. ∑ ((cs) +(kps)
BPO Eligibility Factor is represented as (BPO EF)                +(dk)+(ss) +(dm) +(as))
and KPO Eligibility Factor is represented as (KPO                BPO Attribute Factor is represented as BPO AF and
EF) .SOSS1 represents sum of skill set of BPO EF.                its value is 4. {CS,KPS,DK,SS}
For each attribute the values assigned to them is 1.             KPO Attribute Factor is represented as KPO AF and
SOSS1= Summation of the four attributes they are                 its value is 6.{CS,KPS,DK,SS,DM,AS}
communication skills, knowledge on productivity                  BPO EF=SOSS1/BPO AF



                                                           93                                http://sites.google.com/site/ijcsis/
                                                                                             ISSN 1947-5500
                                                   (IJCSIS) International Journal of Computer Science and Information Security,
                                                   Vol. 9, No. 9, September 2011




KPO EF=SOSS2/KPO AF                                               who cannot specify the exact disease of a patient, a
The values of BPO EF and KPO EF can be illustrated                banker who cannot decide whether to give or not a
through decision tree as depicted below.                          loan for a client, a network administrator who is not
                                                                  able to decide about the exact signature of a given
Decision trees                                                    connection, etc. Hence, in these different examples,
A decision tree is a flow-chart-like tree structure               the expert can provide imprecise or uncertain
allowing to determine the class of an                             classifications expressed in the form of a ranking on
object given known values of its attributes. The                  the possible classes. Ignoring the uncertainty may
visual presentation makes the decision tree                       affect the classification results and even produce
model very easy to understand. It is composed of                  erroneous decisions.        Consequently,     ordinary
three basic elements:                                             classification techniques such as decision trees
1. A decision node specifying the test attribute.                 should be adequately adapted to take care of this
2. An edge corresponding to one of the possible                   problem[5]Decision tree can handle big amounts of
values of the test attribute outcomes. It leads                   data. Their representation of acquired knowledge in
generally to a sub decision tree.                                 tree form is intuitive and generally easy to assimilate
3. A leaf which is also named an answer node,                     by humans [6].Decision tree is popular tool in data
including objects that, typically, belong to the same             mining , it is also well suited for the classification
class, or at least are very similar. For what concerns a          task in that it is reasonably good at a variety of
decision tree, the developer must explain how the tree            classification task.[7,8]
is constructed and how it is used:
• Building the tree: Based on a given training set, a             In Fig 2. The value 2 represent the same branch of
decision tree is built. It consists in selecting for each         KPS. Fig1. Shows the decision tree for the BPO
decision node the appropriate test attribute and also             industry and Fig 2. Shows the decision tree for the
to define the class labeling each leaf.                           KPO industry. By knowing the skill set we can able
• Classification: Once the tree is constructed, it is             to identify the chances of getting into ITES industy.In
used in order to classify a new instance.                         the Decision Tree first we have to check the first skill
We start at the root of the decision tree, we test the            set ,if he posses the skill set then the condition is yes
attribute specified by this node. The                             and we have to check the next skill set. If he does
result of this test allows us to remove down the tree             not possess the required skill then the condition
branch according to the attribute                                 become no then he have to improve(IM) that skill
value of the given instance. This process is repeated             and we can check for the next skill .Likewise we
until a leaf is encountered and                                   have to check all the skill set and find out the
which is characterized by a class.[4]                             eligibility factor. By finding the eligibility factor we
                                                                  can able to say the chances of getting a job in ITES
In many real-world problems, classes of examples in
the training set may be partially defined and even
missing.For example, for some instances, an expert
may be unable to give the exact class value. A doctor




                                                            94                                http://sites.google.com/site/ijcsis/
                                                                                              ISSN 1947-5500
                                                                                        (IJCSIS) International Journal of Computer Science and Information Security,
                                                                                        Vol. 9, No. 9, September 2011




                                SKILLSET




                                  CS

                       y                                         N
                      KPS

                 y                         N                                       IM

                DK
                            N                       IM
       y
                      IM
      SS                                                                                                  KPS              N
y               N                                       DK                                        y
                                                                   N                              DK
BPO
           IM          SS                      y
                                                                                        y                                        IM
                                                                     IM
                 y               N                 SS                                                       IM
                                                                                        SS
                BPO               IM
                                               y                               y
                                                                     SS
                                                                                                      N                             DK                                   DK
                                           BPO
                                                                               BPO                                                          N                                   N
                                                             y            N                  IM              SS              y                                  y
                                                                                                                  N                              IM
                                                             BPO          IM                       y                           SS                                   SS
                                                                                                                                                                                     IM
                                                                                                                                                                               N
                                                                                                                       y                                    y
                                                                                                  BPO             IM                       N
                                                                                                                                                 SS                       IM
                                                                                                                       BPO            IM                                             SS
                                                                                                                                                            BPO
                                                                                                                                            y         N                        y          N          SS

                                                                                                                                           BPO         IM                                 IM              N
                                                                                                                                                                               BPO              y
                                                                                                                                                                                               BPO        IM


                Figure 1. Decision Tree for the Skill Set in BPO




                                                                                                       95                                        http://sites.google.com/site/ijcsis/
                                                                                                                                                 ISSN 1947-5500
                                                                                                         (IJCSIS) International Journal of Computer Science and Information Security,
                                                                                                      Vol. 9,
                                                                                                    SKILLSET No. 9, September 2011



                                                                                                 CS
                                                                                  y
                                                                        KPS
                                                               y                                                        N
                                              y           DK
                      y             SS
                 DM                                                                                                         IM
       y
                               N                                                                    N
y     AS
                                                                         N                                                        2
BPO         N          IM
                                                  N
      IM
                                                                            IM
                                                                                                                   IM
            y    AS

       BPO                                         IM
                      N
                      IM
                                y        DM
             y             AS
                                                                                      y             SS
                                                      N
           BPO                                                     y              DM
                                N             IM
                       IM                                 y        AS                           N
                                                        BPO                               IM
                                                                                                                        N
                           y             AS                               N                                                                                           y
                                                                   IM                                                                                                                 DK
                       BPO                    N                                                                                                   y              SS
                                                                                  AS
                                                                        y
                                              IM
                                                                    BPO                                            IM             y              DM
                                                                                            N
                                                                                       IM
                                                                                                                             y    AS
                                                                                                                                                             N                                  N
                                                                                                y         DM
                                                                                                                            BPO             N          IM                   N
                                                                              y             AS
                                                                                                                                  IM
                                                                                                                    N                                                                                IM
                                                                        BPO
                                                                                                 N            IM                       y         AS

                                                                                                                                      BPO                                       IM
                                                                                          IM                                                                N
                                                                                          y              AS                                           IM
                                                                                                                                                            y         DM                                  y
                                                                                          BPO
                                                                                                                   N                                                                                                        SS
                                                                                                                                             y             AS
                                                                                                              IM                                                                 N         y              DM
                                                                                                                                       BPO
                                                                                                                                                            N              IM
                                                                                                                                                                                      y    AS                       N
                                                                                                                                                       IM
                                                                                                                                                                                                                                           N
                                                                                                                                                                                     BPO                        IM
                                                                                                                                                       y              AS                            N
                                                                                                                                                                                           IM
                                                                                                                                                       BPO                  N                   y         AS
                                                                                                                                                                           IM                                                              IM
                                                                                                                                                                                               BPO
                                                                                                                                                                                                                        N
                                                                                                                                                                                                               IM
                                                                                                                                                                                                                                 DM
                                                                                                                                                                                                                        y
            Figure 2. The Decision Tree based on the Skill Set for KPO                                                                                                                               y              AS
                                                                                                                                                                                                                                           N
                                                                                                                                                                                                BPO
                                                                                                                                                                                                                         N            IM
                                                                                                                                                                                                                IM

                                                                                                                                                                                                                    y            AS

                                                                                                                                                                                                                BPO
                                                                                                                                                                                                                                       N
                                                                                                                                                                                                                                      IM




                                                                                                                        96                                                 http://sites.google.com/site/ijcsis/
                                                                                                                                                                           ISSN 1947-5500
                                                     (IJCSIS) International Journal of Computer Science and Information Security,
                                                     Vol. 9, No. 9, September 2011




                  IV. CONCLUSION                                    [8]. J. Han, M. Kamber, Data Mining: Concepts and
                                                                    Techniques, Morgan Kaufmann, Sanfrancisco 2001.
This Paper aids in improvised decision making for
the computer stream students in choosing their career               AUTHORS PROFILE
path which paves the way to enter ITES industry. At
present students lack in choosing the precise career
path. In this paper ,a decision tree building model
based on the various skill sets possessed by the
students is presented . Firstly, the eligibility factor for
BPO is evaluated by filtering the skill
set(SOSS1).Secondly the eligibility factor for KPO is
evaluated by filtering all the skill sets (SOSS2).The
attributes are analyzed and the correct path is
evaluated. With these factor decision tree is created.
                                                                                     T.Hemalatha, is a research scholar
The results showed that, this method not only
                                                                    in Bharathiar university,Coimbatore , India..She has
improves better decision making but also optimizes                  published papers in international journals.Her area of
the structure of decision tree and gives provision for              interests are Decision Support System, Computer
improving skill sets with alternative options. Due to               Application and Education Technology.
the existence of various skill set possessed by
students, how to choose the vital skill set and
attribute is becoming a difficult task. In addition, the
area of skill set with most reasonable attributes are
worthy of exploration leading to various career path
choice.
                    REFERENCES
                                                                                      Dr. Ananthi Sheshasaayee received
 [1]. Jiang Su and Harry Zhang “A Fast Decision                     her Ph.D in Computer Science from Madras
Tree Learning Algorithm” American Associationa for                  University,India. At present she is working as
Artificial Intelligence 2006.                                       Associate professor and Head, Department of
[2] Sangjae Lee, “Using data envelopment analysis                   computer science, Quaid-e-Millath Government
and decision trees for efficiency analysis                          College for Women, Chennai.She has published 16
andrecommendation of B2C controls” -Decision
                                                                    National and International journals. Her area of
Support Systems 49 (2010) 486–497, ScienceDirect
[3] J. Han, M. Kamber, Data Mining: Concepts and                    interest involve the fields of Computer Applications
Techniques, Morgan Kaufmann, Sanfrancisco 2006                      and Educational technology
[4]Salsabil Trabelsi, Zied Elouedi *, Khaled
Mellouli, “Pruning belief decision tree methods in
averaging and conjunctive approaches” International
Journal of Approximate Reasoning
46 (2007) 568–595, ScienceDirect
 [5] Ilyes Jenhani *, Nahla Ben Amor, Zied Elouedi”
Decision trees as possibilistic classifiers”,
International Journal of Approximate Reasoning
48 (2008) 784–807, ScienceDirect
[6].Han J,Kamber S. Data mining : Conepts and
techniques .Morgan kaufman publishers 2006
[7]. D.Y. Sha . C-H .Liu, “Using Data Mining for
Due Date Assignment in a a Dynamic Job Shop
Environment”, Int J Adv Manuf Technol(2005)
25:1164-1174




                                                              97                                http://sites.google.com/site/ijcsis/
                                                                                                ISSN 1947-5500

				
DOCUMENT INFO
Shared By:
Stats:
views:59
posted:10/12/2011
language:English
pages:7