This chapter provides an introduction to using the Object Constraint Language (OCL) in composing UML models. Our goal in this chapter is to gain an introductory understanding of how to use the OCL.

The Object Constraint Language (OCL)

Requirements management is one of the cornerstone activities in software development. Agile methodologies use dedicated methods, techniques and artifacts in order to implement this activity. Remarkably, Backlog Grooming is the activity of managing and welcoming changing requirements in SCRUM. However, current industrial practices in agile development still tend to render this process in the shape of a list of statements, features and bug fixes that often leads to a blurred view of the goals of the project, the underestimation of client's needs and the decrease of the ability to respond to changes. In this paper we outline an approach that uses goal and agent oriented modelling techniques in order to fill in this "intentional" gap that current industrial approaches lack.

/pdf/stargro-building-i-metrics-for-agile-methodologies-1q6b9mly89.pdf

StarGro: Building i* metrics for agile methodologies

/pdf/melhoria-da-qualidade-para-modelos-orientados-a-objectivos-o-4v3ing4fqk.pdf

Melhoria da qualidade para modelos orientados a objectivos: o caso da abordagem KAOS

Research has shown gender differences in problem-solving, and gender biases in how software supports it. GenderMag has five problem-solving facets related to gender-inclusiveness: motivation for using software, information processing style, computer self-efficacy, attitude towards risk, and ways of learning new technology. Some facet values are more frequent in women, others in men. The role these facets may play when interacting with social goal models is unexplored. We evaluated the impact of different levels of GenderMag facets on creating, modifying, understanding, and reviewing iStar 2.0 models. We performed a quasi-experiment and characterised 180 participants according to each GenderMag facet. Participants performed creation, modification, understanding, and reviewing tasks on iStar 2.0. We measured their accuracy, speed, and ease, using metrics of task success, time, and effort, collected with eye-tracking, EEG and EDA sensors, and participants’ feedback. Although participants with facet levels frequently seen in women had lower speed when compared to those with facet levels more often observed in men, their accuracy was higher. There were also statistically significant differences in visual and mental effort, and stress. Overall, participants were able to create, modify, and understand the models reasonably well, but struggled when reviewing them. Participants with a comprehensive information processing style and a conservative attitude towards risk (characteristics frequently seen in female) solved the tasks with lower speed but higher accuracy. Participants with a selective information processing style (characteristic frequently seen in males) were able to better separate what was relevant from what was not. The complementarity of results suggests there is more gain in leveraging people’s diversity.

Are there gender differences when interacting with social goal models

Goal-oriented requirements engineering approaches provide mechanisms to help eliciting and modelling requirements. This thesis focuses on the i* framework, one of the most important languages for specifying goal-oriented requirements models, which has working groups in over twenty countries and has seven editions of international workshops. There are some problems that can affect the quality of i* models. Their creation can result in complex and/or incomplete models, as well as incorrect ones due to the misuse of the i* notation. These problems contribute to difficulties in the management and understanding of software systems, which in turn leads to increased development costs. The quality of the models should then be a permanent concern, in order to construct models that are correct, complete and whose accidental complexity is minimized. Despite its importance, the effective management of these three aspects is still an open problem. In this thesis, we propose a set of metrics to support the quantitative assessment of the quality attributes complexity, completeness and correctness of i* models. Those metrics are defined through the Goal-Question-Metric approach, being specified both informally, in natural language, as well as formally, with OCL (Object Constraint Language), and incorporated in an i* modelling tool so that they can be automatically collected. The metrics are evaluated both experimentally, through a set of real-world and academic case studies, as well as theoretically, through Weyuker’s properties.

Ana Catarina Gralha de Almeida

Papers

Avaliação da qualidade em modelos de requisitos i

Quality Evaluation of Requirements Models: The Case of Goal Models and Scenarios