Academia.eduAcademia.edu

Figure 9 – uploaded by Ekananta Manalif

See full PDFdownloadDownload figure
‘Table I, Table I], and Table II show that fuzzy- ExCOM risk assessment results are producing a higher correlation with software size and actual effort for all of Figure 9. Project Risk Correlation with Software Size (NASA93 data set)

Figure 9 ‘Table I, Table I], and Table II show that fuzzy- ExCOM risk assessment results are producing a higher correlation with software size and actual effort for all of Figure 9. Project Risk Correlation with Software Size (NASA93 data set)

Related Figures (15)

Risk assessment which is the main activity in a project planning phase, plays a vital role in determining the success of a software development project [9]. However, the success of a risk assessment activity with respect to the popular risk management methods [6][10][11][12][13] is highly dependent on human judgment and experience, therefore risk assessment in software project is perceived as being too effort intensive and costly.
Risk assessment which is the main activity in a project planning phase, plays a vital role in determining the success of a software development project [9]. However, the success of a risk assessment activity with respect to the popular risk management methods [6][10][11][12][13] is highly dependent on human judgment and experience, therefore risk assessment in software project is perceived as being too effort intensive and costly.
Figure 2. Expert COCOMO Risk Taxonomy [4] The risk taxonomy in Expert COCOMO establishes that Software Risk consists of several risks that are related to COCOMO cost factors, such as: Schedule Risk, Product Risk, Platform Risk, Personnel Risk, Process Risk, and Reuse Risk. The software risk taxonomy in Expert COCOMO is described in Figure 2. All risks in Expert COCOMO are defined as the result of a combination of several cost factors. Risk rules determine the level of every risk by mapping 2 cost factors (attributes) according to a risk level assignment matrix as shown in Figure 3.
Figure 2. Expert COCOMO Risk Taxonomy [4] The risk taxonomy in Expert COCOMO establishes that Software Risk consists of several risks that are related to COCOMO cost factors, such as: Schedule Risk, Product Risk, Platform Risk, Personnel Risk, Process Risk, and Reuse Risk. The software risk taxonomy in Expert COCOMO is described in Figure 2. All risks in Expert COCOMO are defined as the result of a combination of several cost factors. Risk rules determine the level of every risk by mapping 2 cost factors (attributes) according to a risk level assignment matrix as shown in Figure 3.
Figure 3. Risk Level Assignment Matrix [4]
Figure 3. Risk Level Assignment Matrix [4]
Figure 6. Fuzzy-ExCOM (fuzzy Expert COCOMO)
Figure 6. Fuzzy-ExCOM (fuzzy Expert COCOMO)
The Fuzzy Inference Engine determines the level of every risk based on the input from 2 cost factors and the risk rules that they represent according to a risk level assignment matrix. Figure 6 shows the risk level matrix for the SCED and CPLX cost factors. Based on this risk-rule, the development project for a product with a very-high complexity (CPLX) level has a very-high scheduling and product risk if it is executed under a very-low project schedule (SCED). Figure 5. Fuzzy-ExCOM (fuzzy Expert COCOMO)
The Fuzzy Inference Engine determines the level of every risk based on the input from 2 cost factors and the risk rules that they represent according to a risk level assignment matrix. Figure 6 shows the risk level matrix for the SCED and CPLX cost factors. Based on this risk-rule, the development project for a product with a very-high complexity (CPLX) level has a very-high scheduling and product risk if it is executed under a very-low project schedule (SCED). Figure 5. Fuzzy-ExCOM (fuzzy Expert COCOMO)
Figure 4. Fuzzy-ExCOM (fuzzy Expert COCOMO)
Figure 4. Fuzzy-ExCOM (fuzzy Expert COCOMO)
The last process in the fuzzy system is the defuzzification process, which produces and translates a fuzzy output to a quantifiable result. The defuzzifier in the fuzzy-ExCOM model translates every fuzzy risk level from an inference process to a quantifiable value that will be used to calculate the project risks based on an equation (1).
The last process in the fuzzy system is the defuzzification process, which produces and translates a fuzzy output to a quantifiable result. The defuzzifier in the fuzzy-ExCOM model translates every fuzzy risk level from an inference process to a quantifiable value that will be used to calculate the project risks based on an equation (1).
TABLE III. RISK CORRELATION WITH SIZE AND ACTUAL EFFORT (INDUSTRY DATA SET)
TABLE III. RISK CORRELATION WITH SIZE AND ACTUAL EFFORT (INDUSTRY DATA SET)
TABLE II. RISK CORRELATION WITH SIZE AND ACTUAL EFFORT (TURKISH DATA SET)
TABLE II. RISK CORRELATION WITH SIZE AND ACTUAL EFFORT (TURKISH DATA SET)
TABLE I. RISK CORRELATION WITH SIZE AND ACTUAL EFFORT (NASA93 DATA SET) Table II shows the correlation results for the TURKISH data set and the correlation results for the INDUSTRY data set is shown in Table III. The correlation chart diagram for risk against software size for NASA93 data set is shown in Figure 9.
TABLE I. RISK CORRELATION WITH SIZE AND ACTUAL EFFORT (NASA93 DATA SET) Table II shows the correlation results for the TURKISH data set and the correlation results for the INDUSTRY data set is shown in Table III. The correlation chart diagram for risk against software size for NASA93 data set is shown in Figure 9.
Figure 8. Fuzzy-ExCOM Model Validation
Figure 8. Fuzzy-ExCOM Model Validation
TABLE V. PROJECT RISK ASSESSMENT USING FUZZY EXCOM (NASA93 PARTIAL DATA)
TABLE V. PROJECT RISK ASSESSMENT USING FUZZY EXCOM (NASA93 PARTIAL DATA)
TABLE IV. PROJECT RISK ASSESSMENT USING EXPERT COCOMO (NASA93 PARTIAL DATA)
TABLE IV. PROJECT RISK ASSESSMENT USING EXPERT COCOMO (NASA93 PARTIAL DATA)
TABLE VII. PROJECT RISK ASSESSMENT USING FUZZY-EXCOM (TURKISH- INDUSTRY DATA)
TABLE VII. PROJECT RISK ASSESSMENT USING FUZZY-EXCOM (TURKISH- INDUSTRY DATA)
TABLE VI. PROJECT RISK ASSESSMENT USING EXPERT COCOMO (TURKISH- INDUSTRY DATA)
TABLE VI. PROJECT RISK ASSESSMENT USING EXPERT COCOMO (TURKISH- INDUSTRY DATA)
Related topics:
Software Engineering
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved