I. Introducción

Studies exhibit that the works focused on software engineering measurement are oriented mainly towards product measurements and in a lower proportion towards the process executed to obtain said product. Measurement of the product is much easier than the measurement of the process and the latter serves to better decision-making about possible changes in business and operating model [1]. According to [2], the monitoring and measurement of the processes are necessary to know if the results are compliance with was planned, including corrective measures. In this sense, the improvement actions require documentation and knowledge of the performance of the processes by defining a set of indicators that supply relevant information, regarding the execution and results of one or several processes in an adequate and representative manner.

Various practices that support the measurement of different entities related to the area of software engineering, have been proposed [3]. Both academics and industry use these practices and empirical studies to generate a variety of proposals aimed at supporting the control of progress in the development of software projects. However, most of these approaches lack theoretical foundations and a common ground that facilitates their implementation, integration, and understanding. Thus, the lack of rigor in most of these approaches encourages the need to introduce languages that assist the stakeholders to compare, evaluate, analyze, simulate, adapt, and compose methods [4].

SEMAT (Software Engineering Method and Theory) supports the representation of modern practices by providing a set of elements to systematically obtain information from the engineering processes and control the activities developed during the software engineering process [5]–[7]. For example, stakeholder management, requirements elicitation, and development of software systems, among others. Such proposal, adopted by the OMG (Object Management Group), was conceived with the purpose of facing a wide variety of problems, including the lack of theoretical bases, lack of consensus, and fashion prevalence the field of software engineering [4]. On the other hand, QFD (Quality Function Deployment) is a method that allows customers to transform qualitative needs into quantitative characteristics, facilitating the deployment of quality functions in subsystems and component parts. Besides, provides a quality framework which considers the assessment and concepts that the client deems appropriate to achieve the satisfaction of their needs, including a sequence of steps and processes for this purpose [8]. QFD is focused on the integration of each of the management processes and the responsibility of each of the stakeholders, obtaining a quality management in all the phases and processes that are involved in the development of a product, including software development [9].

This paper is aimed at providing an approach that satisfy the need of using frameworks standardized and easily adopted to support the measurement in the software development. Particularly, this work is targeted to take advantage of the benefits of QFD to structure a set of indicators for supporting the control of progress in the development of software projects and mainly. These indicators were selected from a literature review and a survey applied to several representative companies related to software development. Then, these indicators are used as input to populate the matrices that make up the QFD quality house. In the same way, we intended to use the capacity provided by the kernel of the essence of SEMAT to supports the representation and theoretical foundation of the developing of software projects. In this sense, an extension to such kernel, related to the measurement of progress, is proposed, and described. This extension is based on the specification provided by QFD, which will be integrated into the framework described by SEMAT.

II. Materials and methods

According to [10], a methodological framework refers to the set of methodological elements, which are required to plan the way in which the development of the research will proceed. Based on the methodological aspects described by [11], a methodological approach for the development of this work is established by following the steps illustrated in Figure 1.

Figure 1.
Steps of the proposed methodology.
the authors based on [11]

Knowledge. In this step, a rigorous search, collection and synthesis of information about SEMAT and QFD was carried out in order to characterize such frameworks and to demonstrate their application in both the academic and business field aimed at supporting the progress measurement in the development of software projects.

Design & elaboration. According to [12], the research design is defined as a general strategy assumed by the researcher to respond to the problem posed. Besides, the design also involves the decisions that are made regarding data collection [13]. In this way, it is necessary to indicate the methodological aspects related to the population and sample that will serve as base in the collection of information.

According to [13], the population and sample establishes the study units as the set of beings that have the characteristic or event to be studied and that are framed in the inclusion criteria that make up the population. In this case, the population refers to those actors who play professional roles in software development, which are related to processes of measuring the state of progress in the development of their software projects and who work in companies dedicated to develop software with national coverage.

Once the population is identified, a representative sample is taken to carry out a survey where the respondents were selected based on their willingness to collaborate, the type of experience, attitude, and necessary knowledge, which resulted in obtaining the maximum benefit from the information collected. A set composed by 11 companies in the software sector was selected. Thus, a technique known as convenience sampling was carried out, which consists of selecting the available cases to which there is access [14]. The selection of the sample was made considering the national and even international coverage of the software projects that are developed by said companies [15]. Data collection techniques were based on direct observation and documentary review supported by instruments such as files, bibliographic reviews, and surveys. The operationalization of the objectives included the establishment of relations between specific objectives and research variables. According to [13], this operationalization process allows the researcher to identify those perceptible aspects of an event that make it possible the recognition of its presence or intensity. For this purpose, tools and resources provided by SEMAT and QFD facilitate the aforementioned operationalization in a clear and precise manner.

Assessment. The assessment of the results of the project was made through a laboratory study intended to validate the answers given by the respondents. This validation was based on the integration of the elements of SEMAT kernel to represent the practice used in the process of measuring the progress status of software projects and the application of the QFD methodology to configure the quality house matrix whose purpose is addressed to identify and analyze the answers given in the survey. Conclusions & recommendations. In this step, the

results obtained and the findings related to measure the state of progress in the development of software projects are presented. In this case, these conclusions and recommendations were based on the characterization of the elements provided by SEMAT kernel and integration of QFD to ensure quality in the processes related to the measurement of the state of progress in the development of projects. Thus, some recommendations are made by providing a diagnosis of the impact of this proposal in relation to the measurement of the state of progress in the development of software projects.

Documentation. The documentation of a research allows the researches to maintain evidence of the activities carried out and achieve their reproducibility, as well as to support the results obtained.

Research presentation. The results in concordance with the objectives of the research are presented. This presentation is intended to demonstrate the degree of impact and importance that this proposal provides regarding to measure the progress status in software projects.

Essence kernel promoted by SEMAT includes a set of elements that are universal to all software engineering efforts. Through the states defined by its elements, Essence kernel [7] provides a novel and effective instrument for reasoning about the progress and health of software development efforts in a method independent manner [5], [6]. The growing number of software projects requires the development of methods that guarantee quality assurance and effective management. Traditionally, proposals to guide changes in project management are based on approaches such as: waterfall model, incremental spiral model, prototype, and rapid application development, among others. In addition, it can be used to increase customer satisfaction by incorporating the voice of the stakeholder in the software development process. Potential defects correspond to the sum of errors found in requirements, design, coding, user documents, and bad corrections. So, the use of QFD to support defect prevention is recommended [16].

In this case we propose a setting of SEMAT according to the specifications required in this study. The elements of SEMAT included in this study are shown in Table 1.

Additionally, QFD is used as a method for developing quality designs to satisfy customer needs by translating consumer desires into key design and quality assurance targets [17]. Voice of Customer (Customer Needs) as input data and information can be obtained from the results of a customer survey conducted. In this case, we use QFD in order to support the identification and analysis of the measures and measure units that were identified by using the mapping review and the surveys practiced to the respondents. The basic steps included by QFD are related to the determination of stakeholder requirements, engineering characteristics, matrix of relationships, roof of the quality house, and target values. However, considering the scope of our research project, we selected the following steps of QFD:

A. Determine stakeholder requirements

Different leaders of companies dedicated to develop software projects were selected as stakeholders who represented the voice of the stakeholder.

B. Determine engineering characteristics

These characteristics correspond to the technical specifications that form the section called “HOW” of the House of Quality described by QFD.

C. Determine matrix of relationships

Based on the information that was registered in the “WHAT” and the “HOW” sections of the House of Quality, the relationships between them are established.

III. Results

SEMAT was used to represent the practice associated with the measurement process of the progress status in software projects. To do this, SEMAT defines three areas of interest involved in the development of a software project: Stakeholder, Solution, and Effort. Once these areas were analyzed, it was established that this research work is based on the area of interest known as Effort. According to [7], this area of interest contains everything that has relation with the team and the way they approach their work. In addition, such area considers those practices aimed at planning, guiding, and monitoring team efforts. In this way, we propose a new practice called “Measure progress status” for supporting the characterization of a specific set of activities and its respective work products. Figure 2 illustrates the representation of this practice with its respective work products. By the other hand, the description of the method to apply QFD in the context of the measurement of the state of progress in the development of software projects is presented. It is important to clarify that the definition of the concepts of measure and measurement are conform to the ISO/IEC/IEEE 15939:2017 standard.

Figure 2.
Practice “Measure progress status”
The authors

In Figure 2, an activity called “Provide measures to show progress and speed” is proposed. This activity is crucial since it involves the essential measures to measure the state of progress in the development of software projects. Therefore, an association with a space activity “Integrate QFD into the measurement process” is established. Thus, QFD is used to support the process by grouping a set of key activities. Such activity space is connected to a set of activities, and these activities are linked to their corresponding competencies. Then, such activities are linked to work products, as illustrated in Figure 3 and Figure 4.

In order to demonstrate the applicability of our proposal, in this section the first three activities related to “Integrate QFD into the measurement process” are executed:

A. Determine stakeholder requirements

Such activity consists of selecting the necessary measures to satisfy the measurement process of the state of progress. Consequently, these measures correspond to the requirements or needs of the stakeholder. These needs arise in a first phase, from a literature review that allowed identifying some common characteristics considered essential to measure the progress of a software project. Based on this information a survey was designed which was applied to the stakeholders selected. In such survey, we asking to stakeholders for their opinion on the measures that they consider relevant to measure the state of progress of software projects. The response options for the respondents corresponded to the various alternatives found during the literature review, and they also had the possibility of adding other alternatives from the point of view of their professional experience. This was done with the purpose of complet-

Figure 3.
Activity space “Integrate QFD into the measurement process”
the authors

Figure 4.
Work products related to activities.
the authors

ing the first qualitative phase based on the analysis of the information and data obtained through the data collection instrument. This information represents the needs from a viewpoint of the clients and is used to create the section called “WHAT” of the House of Quality described by QFD, as illustrated in Table 2.

B. Determine engineering characteristics

This section refers to identify how the needs of the stakeholder can be faced from the view point of the technical experts. In this case, the activity refers to specify the units of measures or technical specifications necessary to satisfy the measurement of the state of progress. So, these units of measurement correspond to the engineering characteristics. Such characteristics, as well as the needs selected in the previous stage, were identified based on review of the literature and later to the elaboration of the survey that was applied to the stakeholders in terms of their opinion about the relevant units of measures to measure the progress status of software projects. These resulting characteristics are illustrated in Table 3.

C. Determine matrix of relationships

It is possible to illustrate a characterization to measure the progress status of software projects. In this case, the needs of the stakeholders and the technical spec-

ifications are considered from two key points of view. The first point corresponds to the measures that are considered relevant for the stakeholders and the second point corresponds to the units of measures to satisfy the needs identified. Figure 5 shows the relationships between the WHAT and the HOW.

After the stakeholder requirements and the engineering characteristics are selected and assigned to the quality house, the next step is to qualify the relationships between them [18]. To do this, it should be noted that the characteristics can affect the requirements in a positive or negative way. Then, a weighting process must be carried out in a joint effort by the roles corresponding to the stakeholders and the technical staff. In this work, the two roles can be performed by the same staff or by different staff, this decision depends on the conditions and characteristics of the development of the software project. This step facilitates the common understanding about the effect of the engineering characteristics on the stakeholder requirements. It is important to mention that the scale to evaluate the answers given according to the level of importance is included in the range from 1 to 5, where each value represents the level of importance to the stakeholders. Such values are: 1 (Not important), 2 (Little important), 3 (Neutral), 4 (Important), and 5 (Very important).

Once the degree of importance of the WHAT’s has been identified, the relationship matrix of the “WHAT” vs “HOW” is calculated. This matrix indicates how much impact the engineering characteristics can have on the stakeholder requirements. Each grid is filled with a score Cij = 0, 1, 3 or 9 to estimate the impact. The relationships between these dimensions are not always 1:1, there may be complex relationships, as well as different levels of relationship. A single “engineering characteristic” can influence several “stakeholder

Figure 5.
Relationships matrix.
the authors

requirements”. This matrix shows the relationships between the “WHAT” and the “HOW”, which are defined considering three levels of relationship: weak relationship, medium relationship, and strong relationship. Based on [19], the values suggested in this process are: 0 (No relationship), 1 (Weak), 3 (Medium), and 9 (Strong). A blank cell indicates that the engineering characteristic has no influence on any stakeholder requirement. This means that the interpretation of “WHAT” and “HOW” was not performed properly. The results obtained from such matrix are shown in Figure 6.

IV. Discussion

The frame provided by SEMAT allows the stakeholders to represent adequately any practice related to software development. In particular, the practice related to measure the progress status during the development of a software project. In this way, we exhibited an approach that support the stakeholders to compare, evaluate, analyze, simulate, adapt, understand, and compose practices in this study area. Then, the integration with a frame aimed at assuring quality in the processes by involving the voice of client was demonstrated. Thus, the collected information was useful for validating the integration and demonstrating the application of the proposal described in this paper. A review more detailed of the results in this study allow to us highlight some aspects. By one hand, according to Table 4, the analysis of the statistical process based on the answers given to the requirements in the section “WHAT” within the quality house demonstrates that the requirements raised by the stakeholders (respondents) have a high-level priority relative to the measurement of the state of progress in software development. An example of this assessment is evidenced

Figure 6.
Relationships “WHAT” vs “HOW”
the authors

in such Table 4, where the degree of importance and quality raised has a rating between “Important” and “Very important”. By other hand, the Figure 6, allow us to identify and make a diagnosis of the degree of importance and degree of impact of the diverse components related to the measurement of the progress status during the development of software projects. This relevance is demonstrated by identifying and relating the requirements of the interested parts and the technical characteristics considered in the development of software projects. These tasks facilitate the establishment of controls to be carried out and decisions making to be made to improve quality and progress level in projects. Figure 7 illustrated a greater degree of impact related to Ndf (number of defects per phase), Chr (amount of real hours), Ttd (Total time of Development peoplemonths), Ea/Ee (Current Effort / Estimated Effort), Nce (number of components with errors), Mf (months per development phase), Che (amount of estimated hours) and Npe (number of peopleyears of effort).

V. Conclusion

This work constitutes a valuable contribution in the field of software engineering in particular related to measuring the progress status in the development of software projects. There is a diversity of measurement approaches, as well as various methods used to develop software projects. Some of these approaches are based on empirical experience and others are framed in globally recognized standards, norms, and frameworks. It was found that not only this variety represents a matter of concern, but also the informal and subjective way in which many of these measurement processes are carried out, cause uncertainty and lack of confidence both in the process and in the results obtained. With the purpose of facing these difficulties, SEMAT was proposed as a representation framework based on a kernel that provides a series of elements leading to the establishment of a common base that facilitates the understanding, analysis, and integration of diverse approaches. To achieve the above, a characterization of the elements provided by the SEMAT that could support the measurement of the progress status of software projects was described. Also, the way in which these characteristics are integrated into the measurement process was described. Additionally, the lack of rigor in the application and configuration of the measurement process, which was evidenced in the review of studies, made it possible to demonstrate the usefulness of integrating QFD (Quality Function Deployment) in this work. The use of QFD for the purpose of supporting the development of software products is widely demonstrated in the literature, however, in this work it was innovatively demonstrated that QFD, can also support the execution of software development processes.

VI. Acknowledgement

This paper presents the results of the research project No. 837 of the Universidad del Quindío entitled “Application of the core of software engineering theory and methods (SEMAT) and the quality function deployment methodology (QFD) to measure the progress state of software projects” and had the financial support of the Vicerrectoría de Investigaciones of the Universidad del Quindío, Colombia.

Referencias

[1] M. Gudelj, “Business Process Management Improvement and a New Model for Performance Management,” in Proceedings on 18th International Conference on Industrial Systems-IS’20, 2022, p.535.

[2] J. Cadena, A. Jiménez, and A. Sánchez, “Nivel de madurez de la gestión de procesos en las medianas empresas de los sectores Metalmecánico y Alimentos Frescos y Procesados en la Provincia de Pichincha (Ecuador),” Revista Espacios, vol. 41, no. 21, pp. 104–122, 2020.

[3] M. J. Simonette, M. E. S. Magalhães, and E. Spina, “PMBOK Five Process Groups and Essence Standard: Perfect Partners?” In 4th International Conference in Software Engineering Research and Innovation (CONISOFT), Puebla, 2016, pp. 53– 58.

[4] F. Ali, A. Raza, M. M. Iqbal, and T. Nazir, “Ontological automation of software essence kernel to assess progress of software project,” Mehran University Research Journal of Engineering & Technology, vol. 41, no. 2, pp. 135–145, 2022.

[5] L. F. Castro, S. Montaño, and E. Espitia, “Chapter # 18. A KAOS representation by using the SEMAT kernel,” Software engineering: methods, modeling, and teaching, vol. 4, pp. 322–337, 2017.

[6] L. F. Castro, E. Espitia, and S. Jaramillo, “Propuesta para evaluar la calidad de proyectos software basada en el núcleo SEMAT y QFD,” in Industria 4.0. Escenarios e impacto, 2018, pp. 57– 70.

[7] OMG, Essence Kernel and Language for Software Engineering Methods (Ver 1.2/2018-10-02). Object Management Group, 2018.

[8] A. Karasan, E. Ilbahar, S. Cebi, and C. Kahraman, Customer-oriented product design using an integrated neutrosophic. AHP & DEMATEL & QFD methodology. Applied Soft Computing, 2022.

[9] A. Ionica, M. Leba and R. Dovleac, “A QFD based model integration in Agile software development,” in 12th Iberian Conference on Information Systems and Technologies (CISTI), Lisboa, 2017, pp. 1–6.

[10] J. Alarcón, La influencia del derecho urbanístico en el desarrollo de la ciudad sostenible. El caso de republica de Ecuador. España, 2021.

[11]C. Arroyave, A. Maya, and C. Orozco, “Aplicación de QFD en el proceso de ingeniería de requisitos,” Tesis, Escuela de ingeniería, Universidad Eafit, 2007.

[12] M. L. Baran, “Mixed methods research design,” in Research Anthology on Innovative Research Methodologies and Utilization Across Multiple Disciplines, IGI Global, 2022, pp. 312–333.

[13] J. Hurtado de Barrera, El proyecto de investigación, Sexta ed. Venezuela: Ediciones Quirol, 2008.

[14] R. Hernández, C. Fernández, and M. Baptista, Metodología de la Investigación. México: McGraw Hill Educación, 2014.

[15] Fedesoft, Estudio de caracterización del sector de teleinformática, software y TI en Colombia. Federación Colombiana de la Industria del Software y Tecnologías Informáticas Relacionadas, Consulted in may 12 / 2022.

[16] M. Fargnoli, N. Haber, D. Platti, and M. Tronci, The soft side of QFD: a comparative study on customer requirements’ prioritization in the food sector, 2021.

[17] F. Fatahillah, E. Rimawan, A. Waskito, and S. Apfiasari, “Strategy to Win the Competition by Developing Products and Services Using the QFD Method in Moldbase Manufacturing,” International Journal of Special Education, vol. 3, no. 3, 2022.

[18] H. Mardhiana, D. Rachmawati, F. D. Winati, and A. Z. Yamani, “Implementation of Quality Function Deployment (QFD) for Decision Making in Improving Integrated Academic Information System,” INTENSIF: Jurnal Ilmiah Penelitian dan Penerapan Teknologi Sistem Informasi, vol. 6, no. 1, pp. 92–107, 2022.

[19] A. Rampal, A. Mehra, R. Singh, A. Yadav, K. Nath, and A. Chauhan, “Kano and QFD analyses for autonomous electric car: Design for enhancing customer contentment,” Materials Today: Proceedings, 2022.