0 оценок0% нашли этот документ полезным (0 голосов)
92 просмотров6 страниц
This article presents an innovative approach to the determination of a core set of Quality Measure Elements (QME), which are critical to the ISOIIEG 25000 SQuaRE series of standards. The research identifies three major opportunities for improvement of the current ISO / IEC 25021 standard: better accessibility for the sofware development industry, consistency through abidance to ISO sofware and system engineering vocabulary and ISO metrology standard.
This article presents an innovative approach to the determination of a core set of Quality Measure Elements (QME), which are critical to the ISOIIEG 25000 SQuaRE series of standards. The research identifies three major opportunities for improvement of the current ISO / IEC 25021 standard: better accessibility for the sofware development industry, consistency through abidance to ISO sofware and system engineering vocabulary and ISO metrology standard.
This article presents an innovative approach to the determination of a core set of Quality Measure Elements (QME), which are critical to the ISOIIEG 25000 SQuaRE series of standards. The research identifies three major opportunities for improvement of the current ISO / IEC 25021 standard: better accessibility for the sofware development industry, consistency through abidance to ISO sofware and system engineering vocabulary and ISO metrology standard.
serIes Dominique St-Louis, Witold Suryn ETS - Ecole de Technologie Superieure I d.stlouis@videotron.caJ witold.suryn@etsmtl.cal Abstract - This article presents an innovative approach to the determination of a core set of Quality Measure Elements (QME), and a comparative analysis aimed at enhancing applicability of ISOII E G 25021, which are critical to the ISOIIEG 25000 SQuaRE series of standards. The research presents the methodology, allowing for moving from an outdated and extensive set of measu res, to a core list of base measu res, which can be further diversified through contexts of use, parameters, and considerably simplify the development of derived measures. Keywords- Sofware Quality Measure Elements, Base Measure, QME Categorization, Knowledge Base Repository. I. INTRODUCTION This research identifes three major opportunities for improvement of the current ISO/IEC 25021 standard: better accessibility for the sofware development industry, consistency through abidance to ISO sofware and system engineering vocabulary [19] and ISO metrology standard [6], and fnally the design of the standard itself. Complexity of ISO/IEC 25021, a similar concern to the one expressed about ISO/IEC 9126, makes using this standard difcult. It is also the reason why sofware quality measures remain a vague or non-existent concept for many companies to this day. Making the quality measurement-related part of ISO/IEC 25000 SQuaRE series of standards readily available, usefl, inexpensive to implement, and easy to use, requires the development of a concise and scientifcally sound set of basic measurement construction modules, as well as Quality Measurement Elements or QMEs. Consistent adherence to the measurement theory [6] would improve the credibility of the whole series of standards, while making it scientifcally correct, increasing its reliability, and enhancing its trustworthiness amongst the scientifc community. As ISO/IEC 25021 makes the basis for the development of several other standards in the SQuaRE series (ISO/IEC 25022-23 replace [2], [3] and [4]) the importance of its correct design is crucial. The objective of the presented research was to determine a diferent (than in ISO/IEC 25021) approach to defning a precise set of base quality measures, which would serve as a basis for further development of the majority of derived quality measures. The results obtained provide such set of base measures that are easy to understand by the sofware industry professionals, compliant with ISO/IEC 15939 standard and propose a design that allows for intuitive construction of derived measures. This paper describes the development of the process allowing for selecting quality measures fom an existing catalogue and the method used to determine a core set of QMEs. Also, the development of a digital knowledge base for sofware quality measures proposed over the course of this research would be one of its most important results, as it potentially leads to opportunities to frther develop the SQuaRE series. II. LITERA TOE RVW Several eXlstmg scientifc articles present critical judgments of ISO/IEC sofware quality and measurement standards. The most renowned analyses are briefy discussed below. ISOIIEG 9126 and its Technical Repors In 1991, ISO/IEC 9126 was published as the frst consensus on the terminology for quality characteristics of sofware evaluation. Between 2001 and 2004, ISO published an expanded version containing the quality model and the measures proposed for this model [8]. The current version of the ISO/IEC 9126 standard [1][2][3][4] consists of one normative reference, the Quality models [1], and three technical reports: External metrics [2], internal metrics [3], and Quality in use metrics [4]. Researchers have found that ISO/IEC 9126 standard has some major faws, even if it seems the best available so far. For one, the terminology of ISO/IEC 9126 is not consistent with the classical terminology of measurement science and engineering, which causes a real concern for the validity of proposed measures. Also, measures in ISO/IEC 9126 in many cases are not identifed and/or classifed as base or derived measures, which stays in confict with ISO/IEC 15939 measurement information model [8]. ISO standards 24765 [19] and 25000 both refer to ISO/IEC 15939 [6] defnition of terms for base and derived measures which are: - Base measure: measure defined in terms of an attrbute and the method for quantifing it. A base measure is functionally independent of other measures. 978-1-4673-2421-2/12/$31.00 2012 IEEE 3120 - Derived measure: measure that is defined as a function of two or more values of base measures. Abran et al. [8] analyzed the ISO/IEC 9126 standard and suggested ways to improve it: The standard should comply with the ISO and general terms of metrology [10][17]. Measures in [1][2][3] and [4] are merged and should be reorganized into base and derived measures to comply with ISO/IEC 15939 [10]. Map the information to the measurement model provided by ISO/IEC 15939 [6]. Measurement results described by ranges such as [0 .. 1] could be presented qualitatively. A score of 100% could be described as excellent, and a score of 80% as very good [11][12]. Other researchers have also demonstrated the impairments in the structure of the ISO/IEC 9126 quality model. lung et al. argued that the use of the six sub-characteristics of Compliance leads to two kinds of results: (I) each compliance sub-characteristic belongs to its characteristic as defned in ISO/IEC 9126, (2) the six compliance sub characteristics form a separate dimension [13]. The results of their research showed an ambiguity in the structures of the characteristics and sub-characteristics in ISO/IEC 9126. As for Abran et al. [15] they suggest reviewing the structure of the Usability model by adding Learabilit and Security, based on their analysis of ISO/IEC 25000 SQuaRE. Chua and Dayson [14] discovered that the Feasibilit characteristic adds specifc factors such as Consistency, Simplicity, Legibilityand Col or use. They also suggest adding the Help sub-characteristic to Usability characteristic and the global characteristic of User satisfaction. In the subject of the implementation of the measures in industry, Boegh [9] indicated that the measures listed in ISO/IEC 9126 could not be applied directly because they entailed a degree of imprecision. In addition, the author mentioned that engineers in the industry should devote considerable efforts to redefne the measures and implement formalized approaches. ISO/IEC 2500 - SQuaRE Pushed by its own pace of evolution, the progresses in information technology [8] and the maturity of the industry, ISO recognized the need of modifcations of the past quality standards. As the result, ISO/IEC SC7 developed a new generation of standards in the sofware quality domain, creating the Sofware Quality Requirements and Evaluation (SQuaRE) series of standards. Being the part of this new series ISO/IEC 25021 has the purpose "to defne and/or design an initial set of QME ... " In its 2007 edition of ISO/IEC 25021 [5] ISO/IEC JTCI SC7 addressed several of the issues identifed in ISO/IEC 9126. For example, the standard is now consistent with the theory of metrology as well as the sofware and system engineering vocabulary. Concerning the reorganisation of the measures, ISO/IEC 25021 has identifed and classifed a set of base measures in addition to the current development of other standards concerning derived measures. Moreover, one of its key goals was the harmonization and coordination with ISO/IEC 15939 [8]. The standard also provides a set of rules for the design of QMEs and for verifcation of the design of existing QMEs. Its content provides the link between ISO/IEC 9126, and the subsequent SQuaRE series. ISO/IEC 2502n-Quality Measurement Division is made of six standards: 25020: Measurement reference model and guide 25021: Quality measure elements .25022: Measurement of quality in use (revision of 9126-4) 25023 :Measurement of system and sofware product quality (revision of 9126-2 & 3) .25024: Measurement of data quality .25025: Measurement of IS service quality III. RESEARCH APPROACH Selection of a Repository of measures The initial research approach was to propose a set of base measure elements that would be at the same time science based (i.e. following metrology principles), usefl and accessible to the sofware industry and utilized as an international reference. The catalog of measures was to be built with all the existing quality measures known in international reference set of ISO/IEC JTCI documents [2][3][4][5] and [7]. Besides of the SQuaRE series of standards no other internationally recognized sofware measurement standards referring to ISO/IEC 15939 principles were found. The measures published in ISO/IEC 9126 standard [2][3][4] were taken as the basis of this research because they make the foundation for the development of the SQuaRE series and at the same time the only available complete set of measures. Due to the extensive list of measures that were found, (including duplicates and synonyms) a turn to a more simplistic approach was taken. Only measures in ISO/IEC 9126 parts 2, 3 and 4 [2][3][4] were considered for the experiment. ISO/IEC 25021 was kept in the repository but only for the fnal research comparison analysis as it was considered the control group, while [2][3][4] made the experiment group. Although ISO/IEC 25021 has already identi fed a set of base measures, thus dramatically reducing the number of QMEs proposed by ISO/IEC 9126, QMEs defned in ISO/IEC 25021 were not used for the experiment. This decision was taken in order to diferentiate the experiment and the control group of QMEs as well as obtaining unbiased experiment results for the selection of base QMEs fom ISO/IEC 9126 repository. Only the fnal results of this research, the core set of QMEs generated by the experiment, was to undergo a comparison analysis with the baseline QMEs fom ISO/IEC 25021. 3121 Selection of a core set of base QMEs The base measure elements, even if they primarily apply to the sofware quality measurement, must be based on the standard metrology for systems engineering and sofware, ISO/lEe 15939. Metrology is the "science of measurement and its applications" [17]. It "includes all theoretical and practical aspects of measurement, regardless of the measurement uncertainty and feld of application ... " [17]. The entire list of measures published in ISO/lEe 9126 parts 2, 3 and 4 were identifed, then stored in a repository, called "QME Reference Base" 9126-2-3-4 [18]. As Abran mentioned in [17] the technical reports of ISO/lEe 9126 contain over 250 derived measures and 80 base measures. Afer the development of the repository [18], the research team identifed 403 measures including derived and base measures, while ISO/lEe 25021 had identifed and defned more than 67 base measures. Of these 403 measures, only eight were clearly identifed as base measures, and fve more were ambiguous. The numbers presented above (403) are unrefned and, as there was no discrimination of the terminology, it included a large number of repetitive measures with no classifcation in term of base or derived measures. The process presented in Figure 2 summarizes the approach used in this research for selecting a core set of base QMEs fom the ISO/lEe 9126 standard. Note that step 6 of the approach was theoretically verifed but not practically experimented due to absence of published data. Base QM E selection (6) Verification and validation of the accuracYI repeatability and repetitivity of each baseline measurement. (5) Testing of each base measure found in step (4) with published data from past or existing project. (4) Each base measure found in step (3) must pass the criteria of measure element and measurement method defined in ISO/lEe 15939. (3) Identify all common measures found in step (2), and distinguish each measure according to its contexts of use. (2) Divide each measure fund in step (1 ) in base elements (must be reduced to the simplest expression) and identif the contexts of use. (1) list all the base measures set out in 9126 to 2, 3 and 4 and identify if the measure is a derived measure i.e. a composite or a base measure. QME selection Choice of storage and analsis tool The creation of the repository, called the QME reference base was developed with the Microsof Excel [18]. The base and derived measures proposed in ISO/lEe 9126 parts 2, 3, and 4, as well as those identifed in ISO/lEe 25021 were all listed in the reference base or repository. The reference base was also populated with more information relevant to inventory. For example, elements of traceability allowing locating the source of the information fom the technical reports [2][3][4] were added. Information ofen referenced was added to the database, thus avoiding searching the paper versions of the standard. To enhance the communication, follow-up among the research team, rationale, questions, comments about classifcation or organisation were recorded in distinctive felds of the base. The repository tool itself played a major part in the success of this experiment. It eased the search and comparison of text defning the QMEs. Although inexpensive, easy to use, readily available, and understood by all team members, the choice of this particular data storage, Excel database, had its limitations. It was not strong enough to make a systematic analysis of the data it contained. A relational database management system (RDBMS) would have allowed for more volume and capacity, supported more complex data structures and processing, data evolution and development. Information in ISO/lEe 9126 could have been superposed to the terminology fom ISO 24765 [19] to assist in the interpretation of the text, the organization and the differentiation of the QMEs and the mapping of the information to the measurement model of ISO/lEe 15939 (see Figure 8), thus improving the analytical capabilities. In essence, the choice of Excel over a RDBMS was motivated strictly by time restriction. Nevertheless, a RDBMS would have been incontestably a superior repository tool option. Core QME determination The measures input into the research reference base were classifed, categorized and sorted according to the vocabulary used in ISO/lEe 9126 standard. Following this step, QME contexts of use were deducted using the metrics information fom ISO/lEe 9126 technical reports. The metrics information encompassed the QME names, purpose, application and measurement method, formula, metric scale, measure type, and target audience. The resulting inventory allowed for determining and contrasting base measures fom derived measures (see example in Figure 3). 1 Technical report Characteristic Data Interpretation 91262 External Attractiveness metrics Sub-characteristic Interface appearance customisability Measure X=A/B X is a derived measure, therefore it must be decomposed to determine the base measures which it is composed of, i. e. A and B A= Number of interface QME base= Number of interface elements customised in elements appearance to user's Context 1= customised in appearance Satisfaction Context 2= to user's Satisfaction B= Number of interface QME base= Number of interface elements that the user elements wishes to customise Context 1= user wishes to customise Fig. 3. Example of determination during preliminary analysis ofa base QM As the next step, harmonization of terms was performed: quotation marks, hyphens and spaces were removed, cases were replaced with upper or lower case when appropriate, but words were not changed. Excluding duplicates and using identical words, the reference base contained: 3122 a set of 121 base measures a set of 170 derived measures Afer these adjustments, the reference base underwent several reviews by the research team, each time, thoroughly examining its information. Below the detailed steps followed during the reference base development. Step 1: Context deteion Afer each metrics was dissected to extract the base QME, it had to be determined if this QME had contexts of use. For example, the QME Number of failure has been identifed as a base measure. Its contexts were numerous; it was referred several times as a data element of metrics found in ISO/IEC 9126. Below, some exemplary contexts for that QME: - Number of failures detected durng defined tral period - Number of failures emerqed afer failure is resolved - Number of failures related to transmission Step 2: Grouping of identical and/or similar QMs Once all contexts identifed and reviewed, identical QME were grouped. This grouping, however, called for some minor QME names adjustment in cases where contexts were similar. For example, Number of message for liD related errors QME and Number of messages for Mean liD fulllment ratio QME were all harmonized with a plural term. A number of words in the reference base were found to use the singular and others the plural form. Afer these initial groupings, there were still several small groups of base QME remaining. Groupings extension was created according to the terminology used in ISO/IEC 9126 standards in order to involve more contexts in the existing QMEs. ISO/IEC 24765 [19] was used as disceming judgment tool as it is the terminological dictionary known and broadly accepted in sofware engineering domain. Step 3: QMs Categorization The grouping easily identifed a trend: some base QMEs were original and other base QMEs, without being secondary, could be extricated fom the original QMEs. Secondary base QMEs could be related to primary base QMEs with reference contexts. A QMEs categorization (see Table 1) was developed by using the same reference base, a feld allowing differentiation of the base QME called "roots" fom the other base QME. The objective was to clearly identif the primary QME on which ISO/IEC 15939 measurement model could be applied. TABLE I QM CATEGORATION QME Description Root base Original Caused by other base QM. They are not derived, but Contextualized base QM whose existence is due to the presence of base other QME (e.g. number of errors is a base QME and number of error messages is also a base QME, but an error is needed to create an error message). Singleton base Li sted on Iy once in the reference base. Obsolete base No longer relevant due to technology evolution. Derived Derived QME Note: At this step of the research, derived QME were not supposed to be found since they had already been broken down into base QM . Step 4: Validation the core with ISOIEC 15939 The previous steps greatly reduced the amount of validation work using the templates of ISO/IEC 15939 by focusing on a small core group of QMEs. Validation work covered options fom selecting Root base QME with the largest number of contexts to creating the Singleton base QME (measures less ofen referred to or sparsely represented). The research team created a visual model prior to the measurement model validation (with ISO/IEC 15939), to better evaluate the work achieved. The visual template that maps the Root base QMEs and their various contexts has been called Contextualized base QME model and is presented in Figure 4. As for Figure 5, it presents a concrete example of the QME number of failures and its different contexts of use as described in ISO 9126-2, 3 and 4. Fig. 4. Base and Contextualized QM model view Encounter during operating concurrently with other software Fig. 5. QM Number of failures and its diferent contexts of use. In the following step, for the validation of these Root base QMEs, measurement models provided by ISO/IEC 15939 were used to prove that the selected base measures met indeed the requirements of this standard. In order to formally defne and assess the Root base QMEs the adjusted template based on 15939 standards was created. The ISO/IEC 15939 template had to be adjusted to the context of this research and since many sections of the original template were linked to derived measures rather than to QME, they were removed, e.g. measurement fnction. 3123 The practical validation of developed QMEs was the experimental creation of several derived measures. Figure 6, presents the outcome for the derived measure A(( Aa1//0/l//, Validation of the F00/0 M/by testing the measures on existing projects in the industry were not completed due to confdentiality agreements with partners and in consequence due to time constraints. Computed confidence limits based on the standard result close to the average _ = .I Access auditability will be achieved. Very wide ` confidence limits suggest a potentially large departure and the need for contingency Fig. 6. QME Number of access and with development of Access auditabilit as a derived measure. SLe 5: Comparison of the developed core set of QMs with ISO/C 25021 The research was concluded by a comparative analysis of the F00/ 0 M/ identifed by this research with those listed in ISO/IEC 25021. The existence of each base QME found by the research was verifed in ISO 25021. When possible, information such as defnition, measurement method, and so on was compared. See Table 2 for details. TALE 2 SOM COMASON ELEMNTS OF TH CORE SET OF QMs Findings Core size Base measure comparison This research ISO 25021 21 measures 67 measures (22 initial measures plus extensions) Number of failures Number of failures Number of system Failures Number of faults (code) Number of faults Number of faults (design) Number of faults (requirements) N um ber of messages Number of errors =Error context Number of error messages Number of messages Number of internal N/A functions Number of external N/A functions "feature" N/A Duration IV. RESULTS The presented research applied a different method of determination of the core set of QMEs than the one used by ISO/IEC 25021 WG6 (Work Group 6). The resulting core set of QMEs is also different fom this proposed in ISO/IEC 25021. Some of the base measures identifed by this research were not presented in ISO/IEC 25021 and vice versa, however, the goal of reducing the core set of QMEs fom 67 to much smaller number of 21 valid QMEs has been achieved. The original measures reference database, which data was extracted fom ISO/IEC 9126-2-3-4, contained 410 lines, one line per measure. By limiting the validation work to F00/ 0 M/ category and QMEs with numerous contexts 14 base QME were identifed (coverage of 245 lines or 60% of the reference base, see Figure 7). Number of To Com b0% Completed external It was estimated that around 70 lines still need to be validated, what explains the 40% fom Figure 7. The 165 remaining lines were rarely referred to by derived measures; they ofen applied to a single derived measure, for example: I/O Utilization, Number of instances of data corruption prevention, Size of sofwa reo Another interesting fnding was that III derived measures could be created fom the 14 identifed QMEs. Considering the total of 178 measures found in ISO/IEC 9126-2-3-4, this gave 62% of coverage (111/178) of all the measures found in this standard. Unfortunately, the coverage ratio could not be verifed against SQuaRE list of derived measures since standards are still under development. V. RECOMNDATIONS A few enhancements could be provided by the 2502n division standards to make its use more widespread. Distribution of 2502n division standards in a digital format, like a knowledge database, would attract a larger population of sofware quality engineers, and increase its use by the industry. Consistent use of ISO/IEC 24765 vocabulary. Systematic use of ISO 15939 measurement templates, providing the mapping of each QME to strengthen the scientifc base and the design of the standard. The standards provided by ISO are prepared on paper or as PDF fles. This is considered as a limitation for users, since they must go through several documents in order to fnd the necessary measure elements for their projects. The recommendation of providing a digital knowledge base (see Figure 8) to the industry is one of the most important fndings of this research. 3124 Fig. 8. QME Knowledge base. ISO 25025 Knowledge base ISOlEC Server Creation Server Tool Server Knowledge base ISOIEC With this recommendation, some practical issues could be addressed: Gather all sofware measurements information on a single storage platform. Provide an embedded process to create new measures, base or derived, associated to a specifc business domain or project, sofware or component. Develop graphical engines to build appropriate measurement models. Provide secure and remote access to all the information, thus, letting ISO be the guardian of the data, while allowing organisations to globally share or protect their own measurement information to other ISO users. Other directions for the refnement of this research are: The application of the same analysis method to the measures listed in ISO/IEC 25021 in order to determine other contexts of measurement since the measures provided in ISO 25021 are also based on ISO 9126-technical reports. Adding new QMEs that refect current development trends such as Agile programming. Although this issue came up as a side effect of this research, it seems logical that the standard evolves in this direction. Removing QMEs that are no longer used due to technology evolution. Adding the defnition of QME context as a parameter complementing measurement method and measurement unit. The 2502n division will surely provide the derived measures, but adding the method linking derived measures and QMEs would supplement the information already in the standards. CONCLUSION To improve the ISO/IEC 2502n division standards, the use of a digital content, as proposed in this paper, would be benefcial to the IT industry. A database support and user fiendly applications that allows quality engineers to quickly navigate and research would be a major improvement to these standards. The tools, such as an automatic calculation of derived measures could also be considered. Furthermore, sofware quality measurement activities could easily be integrated to existing development processes. Hence, making it a great tool for improving sofware quality overall. ACKNOWLEDGMNT This article has been developed in the context of a research project realized at Ecole de Technologie Superieure, Montreal, Canada. The authors would like to thank Guillaume Bellosguardo, Salah-Eddine Benbrahim and Charles Hebert for their contribution to this research. Finally, the authors are grateful to fellow students of class MGL842 (20 II) for their valuable input. REFERENCES [1] ISO/IEC 9126-1-2003: Sofware Engineering - Product Quality - Part 1: Quality Model. Geneva. Switzerland: International Organization for Standardization, 200 I. [2] ISO/IEC TR 9126-2-2003: Sofware Engineering - Product Quality - Part 2: External Metrics, Geneva, Switzerland: International Organization for Standardization, 2003. [3] ISO/IEC TR 9126-3-2003: Sofware Engineering - Product Quality - Part 3: Internal Metrics, Geneva, Switzerland: International Organization for Standardization. [4] ISO/IEC TR 9126-4-2004: Sofware Engineering - Product Quality - Part 4: Quality in Use Metrics, Geneva, Switzerland: International Organization for Standardization. [5] SC7, ISO/IEC DTR 25021-2011: Sofware Engineering - Sofware Product Quality Requirements and Evaluation (SQuaRE) - Quality Measure Elements. [6] ISO/IEC IS 15939-2007: Sofware Engineering - Sofware Measurement Process, Geneva, Switzerland: international Organization for Standardization. [7] IEEE Std. 1061-1998, Standard for a Sofware Quality Metrics Methodology, revision, IEE Standards Dept., Piscataway, N.J., 1998 [8] Abran, R. E. Al-Qutaish, J. M. Desharnais, and N. Habra, "An Information Model for Software Quality Measurement with ISO Standards," in Proceedings of the International Conference on Sofware Development (SWDC-REK), Reykjavik, Iceland, 2005. pp. 104-116. [9] J. Boegh, "Certifying Sofware Component Attributes," IEE Sofware, Vol. 23, No. 3, 2006, pp. 74 - 81. [10] A. Abran, R. E. Al-Qutaish, and l-M. Desharais, "Harmonization issues in the updating of the ISO standards on sofware product quality," Metrics News: Journal of the Sofware Metrics Community, Vol. 10, No. 2, 2005, pp. 35-44. [II] A. Abran, R. E. AI-Qutaish, and l Cuadrado-Gallego, "Analysis of the ISO 9126 on Sofware Product Quality Evaluation from the Metrology and ISO 15939 Perspectives WSEAS Transactions on Computers, Vol. 5, No. 11,2006, pp. 2778-2786. [12] A. Abran, R. E. AI-Qutaish, ad J. Cuadrado-Gallego, "Investigation of the Metrology Concepts in ISO 9126 on Sofware Product Quality Evaluation," Proc. Of the 10th international Conference on Computers (lCComp'06), Vouliagmeni, Athens, Greece, 2006, pp. 864-872. [13] H.-W. Jung, S.-G. Kim, and c.-S. Chung, "Measuring Sofware Product Quality: A Survey of ISO/IEC 9126," IEEE Sofware, Vol. 21, No. 5, 2004, pp. 89-92. [14] B. B. Chua and L. E. Dyson, "Applying the ISO 9126 Model to the Evaluation of an e-Learning System," Proc. Of the 21 st Conference of the Australasian Society for Computers in Learning in Tertiary Education (ASClLITE'04), Perth, Australia, 2004, pp. 184-190. [15] . Abran, A. Khelif, W. Suryn, and A. Sefah, "Consolidating the ISO Usability Models," Proc. Of 11th International Sofware Quality Maagement Conference, Glasgow, Scotland, U, 2003. [16] ISO/IEC, International Vocabulary of Basic and General Terms in Metrology (VIM), Geneva, Switzerland: International Organization for Standardization, 2008. [17] A. Abran, Sofware Metrics and Sofware Metrology, IEEE Computer Society, John Wiley & Sons Inc., New Jersey, 2010. [18] QM Reference Base V2.0, 2011 (not published). [19] ISO/IEC FDIS 24765-2009: Systems and sofware engineering - Vocabulary, Geneva, Switzerland: International Organization for Standardization. 3125