the sixth international conference on elearningfor knowledge-based society 17-18 december 2009,...
TRANSCRIPT
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
Source of Knowledge Blooming Like a LotusKnowledge is the competitive weapon of the 21st century
Intellectual
Professional
Cheerfulness
Morality
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
Forecasting Model for the Students’ Job
Turnover in Thai Industries
Pirapat Chantron
Prasong Praneetpolgrang
Master of Science Program in Information TechnologySripatum University, Bangkok, Thailand
2
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
Background of the Research1
Research Objective2
Theories & Related Research 3
4
Conclusions 5
6
Agenda
Experiments
Future Works
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
4
A number of students transfer their majors of studies or change their majors, drop or resign from the university.
Background of the Research
Many students in the university are not aware whether they should choose to study, any field of studies that match for them in order to work directly with their interests.
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
5
After graduating from the university and get into work, a number of students change their work or resign for the reasons that they cannot find the appropriate or proper work with their major of studies or their interests.
Background of the Research (Cont…)
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
6
These are the reasons that students do not have experience and lack of information in their majors of studies. They unknown individual disciplines well enough, and they found afterward that their studies or their majors and their work didn’t fit with them. It is too late for them to start again.
Background of the Research (Cont…)
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
Research Objective
The purpose of this study is to develop forecasting model for the students’ job turnover in Thai industries.
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
8
Data Mining
Bayesian Networks
Cross-validation
Evaluation
Theories and Related Research
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
Data Mining
9
Data mining technique is based on
statistical analysis, it has been used in finding and describing structural patterns in data segmentation and predictions (Witten and Frank,2005).
This technique has been applied extensively in many industries including banking and finances, education, medical sciences and manufacturing.
Theories
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
Bayesian Networks
10
Specific type of graphical model
which is a directed acyclic graph (Kijsirikul,2003).
All of the edges in the graph are directed and there are no cycles.
Used as a classifier that gives the posterior probability distribution of the class node given the values of other attributes.
Theories (cont.)
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
Bayesian Networks (cont.)
Example of Bayesian Networks
11
Theories (cont.)
C A
B
P(A,B,C) = P(A | B) P(B) P(C | B)
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
Cross-validation
12
Some of the data are removed before training
begins.
When training is done, the data that were removed can be used to test the performance of the learned model.
The Data set is separated into two sets, called the training set and the testing set.
Theories (cont.)
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
13
Correct Percentage =
Number of correct classificationTotal number of classifications
Theories (cont.)
Precision =
Recall =
F-measure =
Number of documents relevant and retrievedTotal number of documents that are retrieved
Number of documents relevant and retrievedTotal number of documents that are relevant
2 x Precision x RecallPrecision + Recall
Evaluation in this System
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
14
Related Research
Research in Data Mining TechniquesResearch Author Year Method
Prediction of Higher Education Students’ Graduation with Bayesian Learning and Data
Mining
Yingkuachat et al 2006 Bayesian Networks
Course Planning of extension education to meet market demand by using data mining techniques-an example of a
university in Taiwan
Hsia et al. 2008 Decision Tree, Association rules, and Decision Forest
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
15
Related Research (cont.)
Research in Data Mining Techniques
Research Author Year
Method
Evaluating Bayesian networks’ precision for detecting students’
learning styles
Garcia et al. 2007 Bayesian Networks
Data Mining Techniques for Developing Education in Faculty of Engineering
Waiyamai et al
2001 association rule,
decision tree
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
16
Student Database
Data Pre-processing
Bayesian Networks
Model
1
2
3
System Framework for the research methodology
Data Pre-processing
Post-processing
Data Mining
Research Experiments
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
17
Data mining techniques (Data Mining) were used in this research to create a relationship model between their majors, having and changing their jobs of persons in public and private organizations by studying from academic performance, profiles, and work background. Data from the total sample set were 2,536.
The table of Krejcie and Morgan was used to define the sample size
Research Experiments (cont.)
Dataset
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
Population Sample size Population Sample size Population Sample size
10 10 45 40 80 66
15 14 50 44 85 70
20 19 55 48 90 73
25 24 60 52 95 76
30 28 65 56 100 80
35 32 70 59 110 86
40 36 75 63 120 92
18
Random Sample Size from the Population which based
on Morgan & Krejcie Table
Research Experiments (cont.)
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
19
n = Sample sizeN = Population sizee = The error of sampling
This study allows the error of sampling on 0.05
Formula,)1( 2Ne
Nn
Research Experiments (cont.)
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
Data were used in this study and the modeling consisted of:
- Information from 6 universities: 3 public and 3 private universities, Kasetsart University. Rajabhat
PranakonUniversity, Rajabhat Lopburi University and private universities
including Sripatum University, Durakit Bundit University and Saint John's
University. - Data from 6 organizations: The CP
(Research and Development), The DTAC, The Department of Transportation, Thai International Airways (Aviation Management), the Department of Cooperative, The Auditing Office and the Office of Bangkhen District Office and The Office of Disease Prevention area 1.
Data were used in this study and the modeling consisted of:
- Information from 6 universities: 3 public and 3 private universities, Kasetsart University. Rajabhat
PranakonUniversity, Rajabhat Lopburi University and private universities
including Sripatum University, Durakit Bundit University and Saint John's
University. - Data from 6 organizations: The CP
(Research and Development), The DTAC, The Department of Transportation, Thai International Airways (Aviation Management), the Department of Cooperative, The Auditing Office and the Office of Bangkhen District Office and The Office of Disease Prevention area 1.
20
Research Experiments (cont.)
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
21
Universities Population (N)
Sampling Size
(n)
Kasetsart University 10,558 385
Phranakhon Rajabhat University
4,358 366
Thepsatri Rajabhat University
1,936 331
Sripatum University 4,820 369
Dhurakij Pundit University
3,400 358
Saint john's University 2,862350
Research Experiments (cont.)
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
22
Company Population (N)
Sampling size(n)
Charoen Pokphand1400 311
Dtac6000 375
Department of Land Transport1370 309
Thai Airways1700 324
Cooperative Auditing Department2400 342
Bangkhen District office850 272
22
Research Experiments (cont.)
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
23
Universities sample
Kasetsart University 237
Phranakhon Rajabhat University
241
Thepsatri Rajabhat University 228
Sripatum University 245
Dhurakij Pundit University 242
Saint john's University 238
Total 1431
Company sample
Charoen Pokphand 270
Dtac 130
Department of Land Transport
190
Thai Airways 252
Cooperative Auditing Department
137
Bangkhen District office 126
Total 1105
23
Data from the total sample set were 2,536
Research Experiments (cont.)
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
AttributeDescripti
on
MatchEdu
Function match with the studying field
BROTHERRank order
in the family
STATUSStudent
status
LOCATION Location
DOMICILE Home town
PARENT_STATUS
Parent status
OCC_FATFather
occupation
OCC_MOTMother
occupation
FAM_INCOMEFamily
income
Work ChangeWork
Changing
Attribute
Description
Gender Gender
Uni Type
University Type
Major Field of Education
GpaLevel
Accumulate Grade point average at the last semester
TimeFindWork
Period of experience
PositionPosition of the
job
CompanyType
Company
Salary Job salary rate
GPA_Old GPA
24
ATTRIBUTE OF DATASET
Research Experiments (cont.)
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
Experimental Results
Research Experiments (cont.)
Work ChangeWork Change
SalarySalaryMajorMajor
PositionPosition
Model of the variable that effect to the work changing.
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
26
== Run information ===Test mode: 10-fold cross-validation=== Classifier model (full training set) ===Naïve Bayes Classifiernot using ADTree=== Summary ===Correctly Classified Instances 2280 97.2634 %Incorrectly Classified Instances 256 2.7366 %Kappa statistic 0.8633Mean absolute error 0.0742Root mean squared error 0.1872Relative absolute error 25.1745 %Root relative squared error 48.9402 %Total Number of Instances 2536.0000
The predicting model for work changing was constructed in order to prove the accuracy of data mining technique by using Bayesian Networks. The result indicated that the accuracy was 97.26%. This study suggests the graduated student to used the factors that effect to his working, those are field of study, Major, Position and Salary. These variables are suitable for model constructing to predict the changing of work opportunity.
Research Experiments (cont.)
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
27
In conclusion, it was found that variables effect the description of the factors affecting the change of the job: major, position of the job and job salary rate.
CONCLUSION
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
Applying data mining technique for prediction. In order to increase the prediction power of classification, alternative feature selection might be applied to select importance attributes before classification.
Increase sampling size in the next research, include universities sampling and organizations in order to develop the model more effectively.
Future works
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
References [1] K. Waiyamai, T. Rakthanmanon and C. ngsiri, “Data Mining Techniques for Developing Education in Engineering Faculty,” NECTEC Technical Journal, volume III, no.11, 2001, pp. 134-142.
[2] B. Kijsirikul, Artificial Intelligence, Department of Computer Engineering, Faculty of Engineering, Chulalongkorn University, 2003.
[3] J. Yingkuachat, B. Kijsirikul and P. Praneetpolgrang, “A Prediction of higher Education Students’ Graduation with Bayesian Learning and Data Mining,” in Research and Innovations for Sustainable Development Conference, 2006.
[4] T. Hsia, A. Shie and L. Chen, “Course Planning of extension education to meet market demand by using data mining techniques-an example of chinkuo technology university in Taiwan,” Expert Systems with Applications, volume 34, Issue 1, 2008, pp. 596-602.
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
References (cont.) [5] I. H. Witten and E. Frank, Data Mining: Practical Machine Learning Tools and Techniques, Second Edition, Morgan Kaufmann, San Francisco, 2005.
[6] WEKA, http://www.cs.waikato.ac.nz/ml/weka, 17 September 2007.
[7] P. Garcia, A. Amandi, S. Schiaffino and M.Campo, “Evaluating Bayesian networks’ precision for detecting students’ learning styles,”Computer & Education, Volume 49, Issue 3, 2007, pp. 794-808.
[8] M. Xenos, “Prediction and assessment of student behaviour in open and distance education in computers using Bayesian networks,” Computer & Education, Volume 43, Issue 4, 2004, pp. 345-359.
The Sixth International Conference on eLearningfor Knowledge-Based Society17-18 December 2009, Srisakdi Charmonman IT Center,Assumption University,Suvarnabhumi Campus, Bangkok Metro, Thailand
M
aste
r of
Sci
ence
Pro
gram
in I
nfo
rmat
ion
Tec
hn
olog
y , S
rip
atu
m U
niv
ersi
ty, B
angk
ok, T
hai
lan
d
Thank You for your kind attentionThank You for your kind attention