Raphael Pereira de Oliveira

A Feature-Driven Requirements Engineering Approach for Software Product Lines

The importance of Requirements Engineering within software development has long been established and recognized by researchers and practitioners. Within Software Product Lines (SPL), this activity is even more critical because it needs to cope with common, variable, and product-specific requirements not only for a single product but for the whole set of products in the family. In this paper, we present a Feature-Driven Requirements Engineering approach (FeDRE) that provides support to the requirements specification of software product lines. The approach follows a strategy where features are realized into functional requirements considering the variability captured in a feature model. It also provides guidelines on how to associate chunks of features from a feature model and to consider them as the context for the Use Case specification. The feasibility of the approach is illustrated in a case study for developing an SPL of mobile applications for emergency notifications. Preliminary evaluations on the perceived ease of use and perceived usefulness of requirements analysts using the approach are also presented.

Evaluating Lehman’s Laws of Software Evolution within Software Product Lines: A Preliminary Empirical Study

The evolution of a single system is a task where we deal with the modification of a single product. Lehman’s laws of software evolution were broadly evaluated within this type of systems and the results shown that these single systems evolve according to his stated laws over time. However, when dealing with Software Product Lines (SPL), we need to deal with the modification of several products which include common, variable and product specific assets. Because of the several assets within SPL, each stated law may have a different behavior for each asset kind. Nonetheless, we do not know if the stated laws are still valid for SPL since they were not yet evaluated in this context. Thus, this paper details an empirical investigation where four of the Lehman’s Laws (LL) of Software Evolution were used in an SPL industrial project to understand how the SPL assets evolve over time. This project relates to an application in the medical domain developed in a medium-size company in Brazil. It contains 45 modules and a total of 70.652 bug requests in the tracking system, gathered along the past 10 years. We employed two techniques - the KPSS Test and linear regression analysis, to assess the relationship between LL and SPL assets. Finally, results showed that three laws were supported based on the data employed (continuous change, increasing complexity, and declining quality). The other law (continuing growth) was partly supported, depending on the SPL evaluated asset (common, variable or product-specific).

Defining and validating a feature-driven requirements engineering approach

The specification of requirements is a key activity for achieving the goals of any software project and it has long been established and recognized by researchers and practitioners. Within Software Product Lines (SPL), this activity is even more critical owing to the need to deal with common, variable, and product-specific requirements, not only for a single product but for the whole set of products. In this paper, we present a Feature-Driven Requirements Engineering approach (FeDRE) that provides support to the requirements specification of SPL. The approach realizes features into functional requirements by considering the variability captured in a feature model. It also provides detailed guidelines on how to associate chunks of features from a feature model and to consider them as the context for the Use Case specification. The evaluation of the approach is illustrated in a case study for developing an SPL of mobile applications for emergency notifications. This case study was applied within 14 subjects, 8 subjects from Universitat Politecnica de Valencia and 6 subjects from Federal University of Bahia. Evaluations concerning the perceived ease of use, perceived usefulness, effectiveness and efficiency as regards requirements analysts using the approach are also presented. The results show that FeDRE was perceived as easy to learn and useful by the participants.

On the Relationship between Features Granularity and Non-conformities in Software Product Lines: An Exploratory Study

Within Software Product Lines (SPL) features are well understood and facilitate the communication among SPL developers and domain experts. However, the feature specification task is usually based on natural language, which can present lack of clarity, non-conformities and defects. In order to understand the feature non-conformity in SPL, this paper presents an empirical study to investigate the possible correlation between feature granularity and feature non-conformity, based on an SPL industrial project in the medical domain. The investigation aims at exploring the features non-conformities and their likely root causes using results from a previous study, which captured and classified 137 features non-conformities, identified in 92 features. The findings indicated that there is significant association between the variables feature interaction and feature granularity. From predictive models to estimate feature non-conformities based on feature granularity and feature interaction values, the variable feature interaction presented positive influence on the feature non-conformity and the variable feature granularity presented negative influence on the variable feature non-conformity.

On the Relationship between Inspection and Evolution in Software Product Lines: An Exploratory Study

Manage the evolution in Software Product Lines (SPL) can bring some benefits such as keep the trace ability between assets in core assets and products, avoid some irregular growth or decrease before it becomes a threat to the system, and also use the products feedback to improve the core asset quality. In order to understand the evolution in SPL, this paper presents an empirical study to investigate evidence between information from features non-conformities and data from corrective maintenance, based on an SPL industrial project in the medical domain. The investigation aims at tracking the features non-conformities and their likely root causes using results from two preliminary studies. The first one captured and classified the features non-conformities from features specification of nine sub-domains and the second one investigated the evolution of SPL assets along the sub-domains development. The study sample was analyzed using statistical techniques, such as Spearman correlation rank and Poisson regression models. The findings indicated that there is significant positive correlation between feature non-conformities and corrective maintenance. Sub-domains with a high number of feature non-conformities had a higher number of corrective maintenance. Moreover, sub-domains qualified as high risk have also positive correlation with corrective maintenance. This correlation allows the building of predictive models to estimate corrective maintenance based on the risk sub-domain attribute values.

Evaluating Lehman’s Laws of software evolution within software product lines industrial projects

The evolution of a single system is a task where we deal with the modification of a single product. Lehman’s Laws of software evolution were broadly evaluated within this type of system and the results shown that these single systems evolve according to his stated laws over time. However, considering Software Product Lines (SPL), we need to deal with the modification of several products which include common, variable, and product specific assets. Because of the several assets within SPL, each stated law may have a different behavior for each asset kind. Nonetheless, we do not know if all of the stated laws are still valid for SPL since they were partially evaluated in this context. Thus, this paper details an empirical investigation where Lehman’s Laws (LL) of Software Evolution were used in two SPL industrial projects to understand how the SPL assets evolve over time. These projects are related to an application in the medical domain and another in the financial domain, developed by medium-size companies in Brazil. They contain a total of 71 modules and a total of 71.442 bug requests in their tracking system, gathered along the total of more than 10 years. We employed two techniques - the KPSS Test and linear regression analysis, to assess the relationship between LL and SPL assets. Results showed that one law was completely supported (conservation of organizational stability) for all assets within both empirical studies. Two laws were partially supported for both studies depending on the asset type (continuous growth and conservation of familiarity). Finally, the remaining laws had differences among their results for all assets (continuous change, increasing complexity, and declining quality).

Requirements Evolution in Software Product Lines: An Empirical Study

The evolution of the requirements specification is a key activity for maintaining the goals of any software project and it has long been established and recognized by researchers and practitioners. Within Software Product Lines (SPL), this evolution is even more critical due to the necessity of dealing with common, variable, and product-specific requirements, not only for a single product but for the whole set of products. In this paper, we present the Feature-Driven Requirements Engineering Evolution (FeDRE2) approach, which provides support to evolve the requirements specification of SPL. The approach evolves the feature model and the use case specification from an SPL according to the users needs. It also provides detailed guidelines on how to evolve chunks of features from a feature model and their related use case specifications. The evaluation of the approach is illustrated in an empirical study for evolving an SPL of mobile applications for emergency notifications. This empirical study was applied within 16 participants, 9 participants from Federal University of Bahia (Brazil) and 7 participants from Polytechnic University of Valencia (Spain). Evaluations concerning the perceived ease of use, perceived usefulness, effectiveness, and efficiency as regards requirements analysts using the approach are also presented. The results have shown that FeDRE2 was perceived as easy to learn and useful by the participants.