Professional Documents
Culture Documents
Abstract
OBJECTIVE The objective of this paper is to
extend the work by Mendes [15], and to compare four
techniques for Web effort estimation to identify which
one provides best prediction accuracy.
METHOD We employed four effort estimation
techniques Bayesian networks (BN), forward
stepwise regression (SWR), case-based reasoning
(CBR) and Classification and regression trees (CART)
to obtain effort estimates. The dataset employed was of
150 Web projects from the Tukutuku dataset.
RESULTS Results showed that predictions
obtained using a BN were significantly superior to
those using other techniques.
CONCLUSIONS A model that incorporates the
uncertainty inherent in effort estimation, can
outperform other commonly used techniques, such as
those used in this study.
1. Introduction
A cornerstone of Web project management is sound
resource estimation, the process by which resources
are estimated and allocated effectively, enabling
projects to be delivered on time and within budget.
Resources are factors, such as cost, effort, quality,
problem size, that have a bearing on a projects
outcome. Within the scope of resource estimation, the
causal relationship between factors is not deterministic
and has an inherently uncertain nature. E.g. assuming
there is a relationship between development effort and
an applications quality, it is not necessarily true that
increased effort will lead to improved quality.
However, as effort increases so does the probability of
improved quality. Resource estimation is a complex
domain where corresponding decisions and predictions
require reasoning with uncertainty.
In Web project management the complete
understanding of which factors affect a projects
http://www.aia-erospace.org/stats/aero_stats/stat08.pdf
http://www.tchidagraphics.com/website_ecommerce.htm
334
Categorical
ProImpr
Categorical
Metrics
Categorical
DevTeam
TeamExp
Ratio
Ratio
TotEff
Ratio
EstEff
Ratio
Accuracy
Categorical
WEB APPLICATION
TypeApp
Categorical
TotWP
Ratio
335
NewWP
TotImg
Ratio
Ratio
NewImg
Fots
Ratio
Ratio
HFotsA
Ratio
Hnew
Ratio
TotHigh
Ratio
FotsA
Ratio
New
Ratio
TotNHigh
Ratio
Description
nlang
DevTeam
TeamExp
TotEff
TotWP
NewWP
TotImg
NewImg
Fots
HFotsA
Hnew
totHigh
FotsA
New
totNHigh
Median
3.00
2.00
3.00
78.00
30.00
14.00
43.50
3.00
0.00
0.00
0.00
1.00
1.00
1.00
4.00
Std.
Dev.
1.58
2.57
2.16
1048.94
209.82
202.78
244.71
141.67
3.64
66.84
5.21
66.59
3.07
4.07
4.98
Min.
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
Max.
8
23
10
5000
2000
1980
1820
1000
19
611
27
611
20
19
35
Level
Enhancement
New
No
Yes
No
Yes
No
Yes
Num. Projects
66
84
53
97
77
73
85
65
% of Projs
62
21.3
8.7
8
# of Projs
93
32
13
12
% Projects
44
56
35.3
64.7
51.3
48.7
56.7
43.3
Total
effort
Quality
delivered
People
quality
Functionality
delivered
Parent node
Child node
336
p( X | E )
p( E | X ) p( X )
p( E )
(1)
where:
x
x
337
Begin
Structural Development
Evaluation
Identify
values/states
Identify
nodes/variables
Identify
relationships
Parameter Estimation
Expert
Elicitation
No
Data?
Yes
Automated
Learning
No
Further
Elicitation
Accept?
Yes
Model Validation
Domain expert
Model
Walkthrough
Data-driven
Predictive
Accuracy
No
Accept?
Yes
Next Stage
338
Mean Effort
5.2
22.9
63.1
314.9
2,238.9
Enhancement
V. Low
0.6
0.3
0.1
Yes
Low Med. High V. High V. Low
0.6 0.3 0.25
0.3
1
0.1
0.25
0.5
0.7
0.5
No
Low Med. High V. High V. Low
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
1
0.2
0.2
1
0.2
0.2
0.2
Low
0.2
0.2
0.2
0.2
0.2
New
Yes
Med. High V. High V. Low
0.2
0.2
0.34
0.2
0.2
0.34
0.2
0.2
1
0.2
0.2
0.34
0.2
0.2
No
Low Med. High V. High
0.2
0.2
0.2
0.2
0.2
0.2
0.2
1
1
0.2
0.2
0.2
0.51
0.1
0.2
0.09
0.1
NewWP
Very Low
Low
Medium
High
Very High
Very Low
0.6
0.4
TotWP
Low
Medium
High
0.5
0.25
0.5
0.5
0.25
Very High
0.2
0.2
0.2
0.2
0.2
TypeProject
Enhancement
0.79
New
0.21
Tables 10, 11 and 12 NPTs for Documented Process, Use Metrics and Process Improvement
Process Improvement
Use Metrics
Yes
No
Documented Process
Yes
No
Yes
No
Yes
No
0.98
0.8
0.5
0.47
0.02
0.2
0.5
0.53
Use Metrics
Process Improvement
Yes
Yes
0.8
No
0.2
No
0.1
0.9
Process Improvement
Yes
0.5
No
0.5
339
Meaning
Natural logarithm of nlang
Natural logarithm of LDevTeam
Natural logarithm of TeamExp
Natural logarithm of TotEff
Natural logarithm of NewWP + 1
Natural logarithm of NewImg + 1
Natural logarithm of Fots + 1
Natural logarithm of totHigh + 1
Natural logarithm of totNHigh + 1
Dummy variable created from TypeProj where
enhancement is coded as 1 and new is coded as 0
Dummy variable created from DocProc where yes
is coded as 1 and no is coded as 0
Dummy variable created from ProImpr where yes
is coded as 1 and no is coded as 0
Dummy variable created from Metrics where no
type is coded as 1 and yes is coded as 0
Coeff.
1.636
.731
.259
.859
-.942
.193
.612
Std.
Error
.293
.119
.068
.162
.208
.052
.192
t
5.575
6.134
3.784
5.294
-4.530
3.723
3.187
p>|t|
.00
.00
.00
.00
.00
.00
.002
TotEfft
(2)
340
node error
SRE = 100 * 1
(3)
341
BN
34.3
27.4
33.3
SWR
94.8
100
6.7
A1
138.1
85.1
13.3
A2
134.7
85.
13.3
A3
203
91.7
13.3
CART
690.4
83.2
20
8. Conclusions
This paper has presented the results of an
investigation where a dataset containing data on 120
Web projects was used to build a Bayesian model, and
the predictions obtained using this model were
compared to those obtained using another three
342
9. Acknowledgements
We thank: Dr. N. Mosley for his comments; the DE
who validated our BN model and all companies that
volunteered data to the Tukutuku database; A/Prof. F.
Ferrucci for the 15 projects volunteered to Tukutuku.
This work is sponsored by the RSNZ, under the
Marsden Fund research grant 06-UOA-201 MIS.
10. References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
343