Ñîâðåìåííûå èíôîðìàöèîííûå òåõíîëîãèè/ 3.Ïðîãðàììíîå îáåñïå÷åíèå

Pylypchuk Kh., Bilas O.

Lviv Polytechnic National University, Ukraine

Analysis and improvement of personal software process risks estimation model

 

Abstract - This research provides analysis and estimate of individual risks, which enable of software quality increase by reducing the impact of risks on the project.

                                                                                                                                                       I.            Introduction

With speed growth of IT industry and emergence of a large number of software systems, the problem of quality is crucial. Products with high quality are more likely to hold on the software market and bring much more profit. Effective control of the software development process at the individual level, during the whole lifecycle should provide quality improvement.

A significant threat on the successful software projects implementation have different types of risks. Risk analysis at the project level is commonly implemented. The use of risk analysis in the individual development process will help identify, assess and prevent the emergence of personal risks.

                                                                                    II.       Identification and assessment of individual risks

There are many definitions of risk as the existence of uncertainty associated with the onset of adverse event, and damage arising from the onset of this event. Risk, by definition, Software Engineering Institute (SEI) - an opportunity of losses [1].

Risk management in project involves forecasting and determining of the risks, quantitative risk assessment, the development of methods and ways to respond to the risks and their control throughout the project life cycle.

For the individual risk management process the following steps could be defined:

1. Identification of individual risks;

2. Risk Assessment;

3. Development of responses to risks;

4. Risks control.

1. Identification of risks is done by the decomposing of various phases the development process: requirements, process, functional area, technical features or phases of implementation. For identification of risks and their root causes, the project should be divided to the level of detail when risks on the lowest level of structure  could be identified.

After analyzing the existing project risks, the following individual risks could be determined:

-         Insufficient level of personal qualification;

-         Development of functionally incorrect program elements;

-         Inability to estimate the amount of individual work and time performance;

-         Improper individual planning;

-         Lack of personal communication;

-         Inability to work under time pressure;

-         Human factors (illness, etc.).

2. Risk Assessment - is a stage of qualitative and quantitative analysis. Its purpose is a measurement of identified risks in the first stage, namely:

-         Calculation of risks and the consequences of them;

-         Risk assessment;

-         Permissible level defining;

-         Account of risks.

Quantitative risk assessment is essentially a supplement to the qualitative assessment. As a result, quantitative risk analysis obtained numerical determination of the size of individual risks and risk of the whole project. Risk can be defined both in absolute and in relative terms. Measuring risk in absolute terms are useful for specific types of losses, and relative - when comparing projected level losses from the real level. At the stage of risk assessment will use two evaluations: qualitative and quantitative.

Qualitative assessment includes:

-         Determining the effect of individual risk (impact level) for the project;

-         Determination of risk priority.

Define the base levels of impact on the project:

TABLE 1

Levels of risk impact on the project budget

Levels of risk impact

Insignificant

Low

Average

High

Significant

0,05

0,1

0,2

0,4

0,8

0,1

0,3

0,5

0,7

0,9

1

2

3

4

5

6

7

8

9

10

 

Levels of impact measured from 1 to 10 are interpreted as follows:

10 - Project defeated;

9 - exceeding the budget by 40% or no investment in terms of 40%;

8 - exceeding the budget by 30% or no investment in terms of 30%;

7 - exceeding the budget by 20% or no investment in terms of 20%;

6 - exceeding the budget by 10% or no investment in terms of 10%;

5 - marginally exceeded the project budget;

4 - significant using the time or funds limits of the project, but within the limits of budget;

3 - average using the time or funds limits of the project;

2 - low number of time or funds limits of project;

1 - no real influence on the project.

 

To determine the priority risks will use the following equation:

 (1)

where IRP - individual risk priority

P - probability of risk

IL - risk impact level on the project.

 

Under the probability of occurrence a random variable as a measure of risk (R) means the probability of loss or shortfall of income compared to the expected option:

     (2)

where X is a random variable loss;

P (x) - probability of loss.

TABLE 2

Empirical scale acceptable impact level of risk

¹

The probability of an undesirable outcome (value of risk)

Graduation of risk

1

0,0 – 0,1

Minimal risk

2

0,1 – 0,3

Low risk

3

0,3 – 0,4

Average risk

4

0,4 – 0,6

High risk

5

0,6 – 0,8

Maximum risk

6

0,8 – 1,0

Critical risk

 

The risks can be sorted by using the risks priorities. Thus, the risks that have very low priority can be removed from further analysis.

It is also necessary to determine the overall risk of the project. It is defined as an average of all the risks with high priority:

      (3)

where SCcommon risk,

IRP - individual risk priority,

Namount of risks.

Quantitative risk assessment involves determining the probability of the risks and impacts of their consequences on the project and make the right decisions.

Quantitative risk assessment for each define as follows:

        (4)

where   - value of  risk,

 probability of risk,

 – value losses of risk.

       Also determine the total value loss for the entire project:

    (5)

where  - total value of risk loss.

3. Development of risk prevention measures. At this point, embodied in practice conducted in the previous steps results of research. They are:

-    design of possible strategies;

-    choice of optimal strategy;

-    implementation of the chosen strategy.

   TABLE 3

Measures for responding on risks

Risk

Measures for preventing of the risks

Insufficient level of personal training

Improve a level of skills by additional courses or performing self-development.

Development of functionally wrong program elements

Increase a level of qualification

Inability to estimate the amount of individual work and runtime

Develop the projects of different types and different complexity.

Incorrect individual planning

Conduct the composition of difficult task to subtasks and performing of subtasks planning.

Disadvantages of personal communications

Improve personal communication quality.

Inability to work under time pressure

Improve the skills of self-organization and planning of the work.

Human factors (illness, etc.).

Prematurely take a vacation.

4. Control of the risks. The purpose of this stage are:

-         permanent tracking of individual risks in the process of implementation for the chosen strategy;

-         accumulation and analysis of information in order to prevent the emergence of new risks;

-         revaluation and adjustment of the risks;

-         operational decisions in case of any deviations from the chosen strategy.

This stage provides a cyclicity of the process of risk management. In each project must present unknown (uncertain) risks whose share in total amount of risk depends on which field  the project is implemented. In order to reverse the results of such risks in the personal reserve, are laid some financial and time resources. But responding to unknown risks only after their appearing associated with a sufficiently large cost. More efficient is an ahead of events, reseiving  the information about the potential emergence of unknown risk in advance, when it is possible to adjust the process without repeating a number of works.

Conclusion

Summarizing the above, it should be emphasized that the main purpose of risk assessment - to give developers the necessary information to make decisions about the feasibility of participation in one or another project and provide measures to protect against possible financial losses. Different methods for individual assessment can be used for quantitative risk analysis. Methods of quantitative assessment are more sophisticated, and factor of uncertainty decrease. Consequently, identification and assessment of individual risks can greatly improve the quality of software by providing market competitiveness.

References

[1] W. Hayes and J. W. Over. The Personal Software Process (PSP): An empirical study of the impact of PSP on individual engineers. Technical Report CMU/SEI-97-TR-001, Software Engineering Institute, Pittsburgh, PA., 1997.

[2] Ray C. Williams, George J. Pandelios, Sandra G. Behrens. 1999. Software Risk Evaluation (SRE) Method Description

[3] Pisarevskyy I.M. Risk Management / Kharkiv: KSAME, 2008.- 124 p.

[4] T. Demark Valsyruya with Bear: Management dashes in the projects of software Development Provision / Demark T., T. Lister - M .: Office. M., 2005. - 190 p.