Miguel Goulão

Objective: To present an overview on the current state of the art concerning metrics-based
quality evaluation of software components and component assemblies.
Method: Comparison of several approaches available in the literature, using a framework
comprising several aspects, such as scope, intent, definition technique, and maturity.
Results: The identification of common shortcomings of current approaches, such as
ambiguity in definition, lack of adequacy of the specifying formalisms and insufficient
validation of current quality models and metrics for software components.
Conclusions: Quality evaluation of components and component-based infrastructures
presents new challenges to the Experimental Software Engineering community.

Reviews on software engineering literature have shown an insufficient experimental validation of claims, when compared to the standard practice in other well-established sciences. Poor validation of software engineering claims increases the risks of introducing changes in the software process of an organization, as the potential benefits assessment is based on hype, rather than on facts. The community lacks highly disseminated experimental best practices. We contribute with a model of the experimental software engineering process that is aligned with recent proposals for best practices in experimental data dissemination. The model can be used in the definition of software engineering experiments and in comparisons among experimental results.

Background: Often, claims in Component-Based Development (CBD) are only sup-
ported by qualitative expert opinion, rather than by quantitative data. This contrasts
with the normal practice in other sciences, where a sound experimental validation of
claims is standard practice. Experimental Software Engineering (ESE) aims to bridge
this gap. Unfortunately, it is common to find experimental validation efforts that are
hard to replicate and compare, to build up the body of knowledge in CBD.

Objectives: In this dissertation our goals are (i) to contribute to evolution of ESE, in
what concerns the replicability and comparability of experimental work, and (ii) to ap-
ply our proposals to CBD, thus contributing to its deeper and sounder understanding.

Techniques: We propose a process model for ESE, aligned with current experimen-
tal best practices, and combine this model with a measurement technique called
Ontology-Driven Measurement (ODM). ODM is aimed at improving the state of prac-
tice in metrics definition and collection, by making metrics definitions formal and ex-
ecutable, without sacrificing their usability. ODM uses standard technologies that can
be well adapted to current integrated development environments.

Results: Our contributions include the definition and preliminary validation of a pro-
cess model for ESE and the proposal of ODM for supporting metrics definition and
collection in the context of CBD. We use both the process model and ODM to perform
a series experimental works in CBD, including the cross-validation of a component
metrics set for JavaBeans, a case study on the influence of practitioners expertise in
a sub-process of component development (component code inspections), and an ob-
servational study on reusability patterns of pluggable components (Eclipse plug-ins).
These experimental works implied proposing, adapting, or selecting adequate ontolo-
gies, as well as the formal definition of metrics upon each of those ontologies.

Limitations: Although our experimental work covers a variety of component models
and, orthogonally, both process and product, the plethora of opportunities for using
our quantitative approach to CBD is far from exhausted.

Conclusions: The main contribution of this dissertation is the illustration, through
practical examples, of how we can combine our experimental process model with
ODM to support the experimental validation of claims in the context of CBD, in a re-
peatable and comparableway. In addition, the techniques proposed in this dissertation
are generic and can be applied to other software development paradigms.

Domain Specic Languages (DSLs) can contribute to increment productivity, while reducing the required maintenance and programming expertise. We hypothesize that Software Languages Engineering (SLE) developers consistently skip, or relax, Language Evaluation. Based on the experience of engineering other types of software products, we assume that this may potentially lead to the deployment of inadequate languages. The fact that the languages already deal with concepts from the problem domain, and not the solution domain, is not enough to validate several issues at stake, such as its expressiveness, usability,
effectiveness, maintainability, or even the domain expert's productivity while using them. We present a systematic review on articles published in top ranked venues, from 2001 to 2008, which report DSLs' construction, to characterize the common practice. This work conrms our initial hypothesis and lays the ground for the discussion on how to include a systematic approach to DSL evaluation in the SLE process.

Background: Aspect-oriented programming (AOP) is an
emerging programming paradigm whose focus is about improving
modularity, with an emphasis on the modularization of
crosscutting concerns.

Objective: The goal of this paper is to assess the extent to which
an AOP language – ObjectTeams/Java (OT/J) – improves the
modularity of a software system. This improvement has been
claimed but, to the best of our knowledge, this paper is the first
attempting to present quantitative evidence of it.

Method: We compare functionally-equivalent implementations of
the Gang-of-Four design patterns, developed in Java and OT/J,
using software metrics.

Results: The results of our comparison support the modularity
improvement claims made in the literature. For six of the seven
metrics used, the OT/J versions of the patterns obtained
significantly better results.

Limitations: This work uses a set of metrics originally defined
for object-oriented (OO) systems. It may be the case that the
metrics are biased, in that they were created in the context of OO
programming (OOP), before the advent of AOP. We consider this
comparison a stepping stone as, ultimately, we plan to assess the
modularity improvements with paradigm independent metrics,
which will conceivably eliminate the bias. Each individual
example from the sample used in this paper is small. In future, we
plan to replicate this experiment using larger systems, where the
benefits of AOP may be more noticeable.

Conclusion: This work contributes with evidence to fill gaps in
the body of quantitative results supporting alleged benefits to
software modularity brought by AOP languages, namely OT/J.

Models, with different levels of detail, share similar abstractions that can be reused by means of model-driven techniques such as transformations. For example, scenarios are a well-known technique in requirements engineering to represent behavioral flows in a software system. When using UML, scenarios are typically represented with activity models in the early stages of software development, while sequence models are used to describe more detailed object interactions as modeling progresses. This paper defines transformation rules to automate the migration from activity to sequence models. We present a case study illustrating the application of our transformation rules. Our preliminary assessment of the impact of the benefits of using these transformations points to: (i) a reduction of around 50% in the effort building sequence models, (ii) increased trace ability among models, and (iii) error prevention when migrating from different scenario notations.

Domain Specific Languages (DSLs) are claimed to increase productivity, while reducing the required maintenance and programming expertise. In this context, DSL usability by domain experts is a key factor for its successful adoption. Evidence that support those improvement claims is mostly anecdotal. Our systematic literature review showed that a usability evaluation was often skipped, relaxed, or at least omitted from papers reporting the development of DSLs. The few exceptions mostly take place at the end of the development process where fixing problems identified is too expensive. We argue that a systematic approach based on User Interface experimental validation techniques should be used to assess the impact of the new DSLs. The rationale is that assessing important and specially tailored usability attributes for DSLs early in language construction will ultimately foster a higher productivity of the DSL users. This paper, besides discussing the quality criteria, proposes a development and evaluation process that can be used to achieve usable DSLs in a better way.

Domain Specific Languages (DSLs) are claimed to increment productivity, while reducing the required maintenance and programming expertise. In this context, DSLs usability is a key factor for its successful adoption.

In this paper, we propose a systematic approach based on User Interfaces Experimental validation techniques to assess the impact of the introduction of DSLs on the productivity of domain experts. To illustrate this evaluation approach we present a case study of a DSL for High Energy Physics (HEP).

The DSL on this case study, called Pheasant (PHysicist’s EAsy Analysis Tool), is assessed in contrast with a pre-existing baseline, using General Purpose Languages (GPLs) such as C++. The comparison combines quantitative and qualitative data, collected with users from a real-world setting. Our assessment includes Physicists with programming experience with two profiles; ones with no experience with the previous framework used in the project and other experienced.

This work’s contribution highlights the problem of the absence of systematic approaches for experimental validation of DSLs. It also illustrates how an experimental approach can be used in the context of a DSL evaluation during the Software Languages Engineering activity, with respect to its impact on effectiveness and efficiency.

Scenario modeling can be realized through different perspectives. In UML, scenarios are often modeled with activity models, in an early stage of development. Later, sequence diagrams are used to detail object interactions. The migration from activity diagrams to sequence diagrams is a repetitive and error-prone task. Model-Driven Development (MDD) can help streamlining this process, through transformation rules. Since the information in the activity model is insufficient to generate the corresponding complete sequence model, manual refinements are required. Our goal is to compare the relative effort of building the sequence diagrams manually with that of building them semi-automatically. Our results show a decrease in the number of operations required to build and refine the sequence model of approximately 64% when using MDD, when compared to the manual approach.

KAOS is one of the most well-known goal-oriented requirements engineering approaches. Nevertheless, building large KAOS models sometimes results in incomplete and/or complex requirements models that are difficult to understand and maintain. These shortcomings often lead to an increase in costs of product development and evolution. Therefore, for large-scale systems, the ability to manage the complexity and completeness of KAOS models is essential. In this paper, we propose a metrics suite for supporting the quantitative assessment of KAOS models complexity and completeness, in order to support their early identification. We apply the metrics to an example taken from a health club system specification.

Domain Specic Languages (DSLs) are claimed to contribute
to increment productivity, while reducing the required maintenance and
programming expertise. In this context, the usability of these languages
becomes a major issue: if the language is not easy to learn and use, it is
unlikely to be successfully adopted.

It is important to foster high quality DSLs during its engineering process.
We argue that a systematic approach based on User Interface Experi-
mental validation techniques should be used to assess the impact of the
introduction of DSLs in the productivity of DSL user. Productivity can
be fostered by assessing important usability attributes early in the lan-
guage construction . This work's contribution, besides highlighting the
problem of the absence of systematic approaches for experimental vali-
dation of DSLs in general, is to identify existing evaluation approaches
that can be adapted from the eld of User Interfaces.

We can regard Domain-Specific Languages (DSLs) as User Interfaces (UIs) because they bridge the gap
between the domain experts and the computation platforms. Usability of DSLs by domain experts is a key
factor for their successful adoption. The few reports supporting improvement claims are persuasive, but
mostly anecdotal. Systematic literature reviews show that evidences on the effects of the introduction of
DSLs are actually very scarce. In particular, the evaluation of usability is often skipped, relaxed, or at
least omitted from papers reporting the development of DSLs. The few exceptions mostly take place at
the end of the development process, when fixing problems is already too expensive. A systematic
approach, based on techniques for the experimental evaluation of UIs, should be used to assess suitability
of new DSLs. This chapter presents a general experimental evaluation model, tailored for DSLs’
experimental evaluation, and instantiates it in several DSL’s evaluation examples.

Goal-Oriented Requirements Engineering (GORE) has received increasing attention over the past few years.
There are several goal-oriented approaches, each one using different kinds of models. We argue that it would be useful to relate them or even perform transformations among them automatically, in order to understand their similarities and differences, their advantages and disadvantages, allowing a possible migration or comparison between approaches. This is something that has not received enough attention. In this paper
we propose the definition and implementation of goal model transformations between i* and KAOS. As an immediate contribution, the approach can be used to migrate from one goal model to another through automatic model transformations. This approach also contributes to relate the concepts of i* and KAOS models and will help, for example, a development team in making the decision on which approach to follow, according to the nature of the project and the expressiveness of an approach to represent certain concepts
(e.g., obstacles are represented explicitly in KAOS, but not in i*). Another contribution is to facilitate communication among members of the same team, if they are specialized in different approaches.

For years the development of software artifacts was the sole domain of developers and project
managers. However, experience has taught us that the Users play a very important role in
software development and construction. On Domain Specific Languages the inclusion of the
domain experts directly in the development cycle is a very important characteristic, as they have
often an important role in making and constraining the domain of the language.
DSLs are credited with increased productivity and ease of use, but this fact is hardly ever proven.
Moreover, usability tests are frequently only performed at the final stages of the project when
changes have a significant impact on the budget. To help prevent this, in this paper we present a
pattern language for evaluating the usability of DSLs. Our patterns can help show how to use an
iterative usability validation development strategy to produce DSLs that can achieve a high
degree of usability.

Domain-Specific Languages (DSLs) are claimed to bring important productivity improvements to developers,
when compared to General-Purpose Languages (GPLs). The increased Usability is regarded as one of the key benefits of DSLs when compared to GPLs, and has an important impact on the achieved productivity of the DSL users. So, it is essential to build in good usability while developing the DSL. The purpose of this proposal is to contribute to the systematic activity of Software Language Engineering by focusing on the
issue of the Usability evaluation of DSLs. Usability evaluation is often skipped, relaxed, or at least omitted from papers reporting development of DSLs. We argue that a systematic approach based on User Interface experimental validation techniques should be used to assess the impact of new DSLs. For that purpose, we propose to merge common Usability evaluation processes with the DSL development process. In order to provide reliable metrics and tools we should reuse and identify good practices that exist in Human-Computer
Interaction community.

Prediction models of software change requests are useful for supporting rational and timely resource
allocation to the evolution process. In this paper we use a time series forecasting model to predict software
maintenance and evolution requests in an open source software project (Eclipse), as an example of projects with seasonal release cycles. We build an ARIMA model based on data collected from Eclipse’s change request tracking system since the project’s start. A change request may refer to defects found in the software, but also to suggested improvements in the system under scrutiny. Our model includes the identification of seasonal patterns and tendencies, and is validated through the forecast of the change requests evolution for the next 12 months. The usage of seasonal information significantly improves the estimation ability of this model, when compared to other ARIMA models found in the literature, and does so for a much longer estimation period. Being able to accurately forecast the change requests’ evolution over a fairly long time period is an important ability for enabling adequate process control in maintenance activities, and facilitates
effort estimation and timely resources allocation. The approach presented in this paper is suitable for projects with a relatively long history, as the model building process relies on historic data.

Complex software systems inherently require a variety of models used in all of the development stages. A general concern is to guarantee consistency and traceability among these models. Model-driven development (MDD) can help tackle this concern. Although MDD has been mainly used in later development stages, it is relatively unexplored in requirements engineering. In this article, the authors discuss how to
leverage MDD to support consistency and traceability in requirements modeling. To illustrate this, they apply MDD to goaloriented requirements engineering (GORE) by making bidirectional mappings between two well-known GORE approaches (i* and KAOS). The result is an interoperable framework that can be used to migrate from one goal model to another through automatic model transformations, keeping consistency and traceability, so requirements engineers can make the best use of each approach.

A rapidly changing market leads to software systems with highly volatile requirements. In many cases, new demands in software can often be met by extending the functionality of systems already in operation. By modularizing volatile requirements that can be altered at the client’s initiative or according to market demands, we can build a stepping-stone for management of requirements change. The volatility must be managed in a way that reduces the time and costs associated with updating a system to meet the new requirements. In this chapter, we present an approach for handling volatile concerns during early life cycle software modeling. The key insight is that techniques for aspect-oriented software development can be applied to modularize volatility and to weave volatile concerns into the base software artifacts.

There are several tools currently available in the i * community. These tools have different features and purposes. Choosing the most adequate tool for a specific modelling situation can be a challenge. To overcome this diffculty, we present a systematic comparison of the i * tools listed in the i * wiki page, according to their features, syntax coverage and semantic analysis support. Our comparison highlights the
different strengths of those tools, to help identifying situations for which each tool might be particularly useful. We contribute with an aggregated vision of current i * tool support to the body of knowledge of the i * community. In addition, this comparison also helps identifying opportunities for further evolution of the surveyed tools.

Goal-Oriented Requirements Engineering (GORE) approaches have been developed to facilitate the requirements engineers work by, for example, providing abstraction mechanisms to help eliciting and modeling requirements. One of the well-established GORE approaches is KAOS. Nevertheless, in large-scale systems building KAOS models may result in incomplete and/or complex goal models, which are difficult to understand and change. This may lead to an increase in costs of product development and evolution. Thus, for large-scale systems, the effective management of complexity and completeness of goal
models is vital. In this paper, we propose a metrics framework for supporting the quantitative assessment of complexity and completeness of KAOS goal models. Those metrics are formally specified, implemented and incorporated in a KAOS modeling tool. We validate the metrics with a set of real-world case studies and discuss the identified recurring modeling practices.