|
2004 - National Statistics Institute
STATISTICS METHOD USING FOR SOFTWARE QUALITY EVALUATION - 4th november 2004
The approach refers the following:
- target group;
- program homogenous sets;
- program quality characteristics;
- software metrics;
- quality characteristic choosing;
- quality analysis of the software metrics;
- program comparative analysis;
- programmer�s productivity.
Quality aspects of the software production process are based on an important volume of data,
using of some statistics techniques and methods that permit the obtaining of satisfactory results
concerning the estimating of necessary efforts to develop program products and, also, the
efficiency level of the programs at the users.
The all approaching has at its basis manipulating practical procedures of the data sets, the results
obtained from initial data aggregation and the hierarchy of the programs differentiated on many
criteria.
These results are obtained after a research activity developed along many years. It has already
been a speciality literature in this area and many research projects in which the authors have
participated. This activity permitted advanced hypothesis verification for each of problem.
Ion Ivan
Catalin Boja
INFORMATICS APPLICATIONS QUALITY MANAGEMENT- 4 march 2004
Prof. univ. dr. I. IVAN
Prof. univ. dr. I. COSTAS
Dr. Gh. NOSCA
Dr. L. TEODORESCU
QUALITY DEFINITION
- Quality is a fuzzy and complex concept
- Quality approaches:
- transcendental quality is a un-analyzable property that is recognized only by personal experience;
- from the users point of view quality is preference satisfying;
- product-based quality is a measurable variable given by product characteristics;
- production-based quality is avoidance of the cost for product repairing;
- value-based quality is performance to the acceptable costs or prices.
- Users appreciate the product quality through degree of this one to accomplish the
tasks, and the manager of the company who uses the product regards the quality through
the obtained benefits. The Persons who assure the maintenance evaluate the quality through the system complexity and technical documentation.
INFORMATICS APPLICATION
The success of Informatics Applications is based on the degree in which the results
satisfy the client requirements (the results are complete, correct, obtained on schedule, in the asked form and with the asked precision). For a software product that satisfies all the requirements in order to obtain the expected results, the users of this one are obligated to assure the requirements regarding the input data.
DATA QUALITY
The studies demonstrated that, in order to define the data quality, the data consumers uses more dimensions: accuracy, credibility, relevance, opportunity and completeness.
- accuracy/exactitude degree in which the data are correct, reliable and certificated as less errors;
- credibility - degree in which the data are accepted as true and real data
- relevace degree in which the data are applicable and useful for the current activities
- opportunity - degree in which the data age is adequate for current activities
- completeness degree in which data cover the area of the current activities
quantitatively and qualitatively.
SOFTWARE QUALITY
Software Quality the technical, economic, functional and social characteristics
qs = f(qpj, qpr)
- qpj project quality
- qpr process quality
- qs product quality
QUALITY CHARACTERISTICS
Internal Characteristics are measured through product examination.
- Maintenance
- Portability
- Reusability
- Testability
- Understanding facility
- Modifiability
External Characteristics are measured in regarding with the way in which the product respond to the stimulation from the environment. These characteristics affect the product value at the user directly.
- Correctness
- Reliability
- Robustness
- Accessibility
- Interoperability
- Accuracy
- Auto-documentation
- The ability to be structure
The internal characteristics determine the quality of its external characteristics.
GLOBAL QUALITY METRICS
- Total number of defects
- Number of defect in each routine
- Average of the defects on thousands of code lines
- Average time between two fails
- Number of detected compiling errors
APPLICATION QUALITY
- Quality characteristics:
- Robustness
- Errors allowable variation
- Correctness
- Completeness
- Reliability
- Maintenance
qa = F(qa, qd, qr)
qa - application quality
qd - data quality
qs - software quality
qr - result quality
QUALITY PLANNING
Quality planning the activities for the establishing the objectives and the requirements regarding the quality and the requirements regarding the implementation of the quality system elements [ISO 8402] .
Quality assurance systematic planning of the actions needed for quality performing.
Quality control operational activities used in the entire life cycle of Informatics Applications in order to accomplish the quality requirements.
INFORMATICS APPLICATIONS QUALITY MANAGEMENT
The specialists in the quality field say that 85% from quality problems are management problems.
Total quality management orients the Informatics Application production toward the users needs.
Total quality management confers importance to the quality face to the imediat profit. In order to obtain new users it must perform advantages through quality for these ones. The active participative management supposes constructive solutions and the allocation of the resources that contribute to the quality obtaining.
The Informatics Application total quality management is a mixture of administration and coordination, work teams, methods and internal work environment.
The good things are obtained if they are planned, the bad things come by themselves
O. ONICESCU
in work..........
|