Juliana Alves Pereira

FeatureIDE: Scalable Product Configuration of Variable Systems

In the last decades, variability management for similar products is one of the main challenges in software systems. In this context, feature models are used to describe the dependencies between reusable common and variable artifacts, called features. However, for large feature models it is a complex task to find a valid feature combination as product configuration. Our Eclipse plug-in FeatureIDE provides several mechanisms, such as information hiding and decision propagation, which support the configuration process to combine the reusable artifacts in various manners. We illustrate the applications of these mechanisms from a users point of view.

A Systematic Literature Review of Software Product Line Management Tools

Software Product Line (SPL) management is a key activity for software product line engineering. The idea behind SPL management is to focus on artifacts that are shared in order to support software reuse and adaptation. Gains are expected in terms of time to market, consistency across products, costs reduction, better flexibility, and better management of change requirements. In this context, there are many available options of SPL variability management tools. This paper presents and discusses the findings from a Systematic Literature Review (SLR) of SPL management tools. Our research method aimed at analyzing the available literature on SPL management tools and the involved experts in the field. This review provides insights (i) to support companies interested to choose a tool for SPL variability management that best fits their needs; (ii) to point out attributes and requirements relevant to those interested in developing new tools; and (iii) to help the improvement of the tools already available. As a direct result of this SLR, we identify gaps, such as the lack of industrial support during product configuration.

On the Effectiveness of Concern Metrics to Detect Code Smells: An Empirical Study

Traditional software metrics have been used to evaluate the maintainability of software programs by supporting the identification of code smells. Recently, concern metrics have also been proposed with this purpose. While traditional metrics quantify properties of software modules, concern metrics quantify concern properties, such as scattering and tangling. Despite being increasingly used in empirical studies, there is a lack of empirical knowledge about the effectiveness of concern metrics to detect code smells. This paper reports the results of an empirical study to investigate whether concern metrics can be useful indicators of three code smells, namely Divergent Change, Shotgun Surgery, and God Class. In this study, 54 subjects from two different institutions have analyzed traditional and concern metrics aiming to detect instances of these code smells in two information systems. The study results indicate that, in general, concern metrics support developers detecting code smells. In particular, we observed that (i) the time spent in code smell detection is more relevant than the developers’ expertise; (ii) concern metrics are clearly useful to detect Divergent Change and God Class; and (iii) the concern metric Number of Concerns per Component is a reliable indicator of Divergent Change.

A feature-based personalized recommender system for product-line configuration

Today’s competitive marketplace requires the industry to understand unique and particular needs of their customers. Product line practices enable companies to create individual products for every customer by providing an interdependent set of features. Users configure personalized products by consecutively selecting desired features based on their individual needs. However, as most features are interdependent, users must understand the impact of their gradual selections in order to make valid decisions. Thus, especially when dealing with large feature models, specialized assistance is needed to guide the users in configuring their product. Recently, recommender systems have proved to be an appropriate mean to assist users in finding information and making decisions. In this paper, we propose an advanced feature recommender system that provides personalized recommendations to users. In detail, we offer four main contributions: (i) We provide a recommender system that suggests relevant features to ease the decision-making process. (ii) Based on this system, we provide visual support to users that guides them through the decision-making process and allows them to focus on valid and relevant parts of the configuration space. (iii) We provide an interactive open-source configurator tool encompassing all those features. (iv) In order to demonstrate the performance of our approach, we compare three different recommender algorithms in two real case studies derived from business experience.

Software Variability Management: An Exploratory Study with Two Feature Modeling Tools

Software Product Line (SPL) is becoming widely adopted in industry due to its capability of minimizing costs and improving quality of software systems through systematic reuse of software artifacts. An SPL is a set of software systems sharing a common, managed set of features that satisfies the specific needs of a particular market segment. A feature represents an increment in functionality relevant to some stakeholders. There are several tools to support variability management by modeling features in SPL. However, it is hard for a developer to choose the most appropriate feature modeling tool due to the several options available. This paper presents the results of an exploratory study aiming to support SPL engineers choosing the feature modeling tool that best fits their needs. This exploratory study compares and analyzes two feature modeling tools, namely FeatureIDE and SPLOT, based on data from 56 participants that used the analyzed tools. In this study, we performed a four-dimension qualitative analysis with respect to common functionalities provided by feature modeling tools: (i) Feature Model Editor, (ii) Automated Analysis of Feature Models, (iii) Product Configuration, and (iv) Tool Notation. The main issues we observed in SPLOT are related to its interface. FeatureIDE, on the other hand, revealed some constraints when creating feature models.

Quantitative and Qualitative Empirical Analysis of Three Feature Modeling Tools

During the last couple of decades, feature modeling tools have played a significant role in the improvement of software productivity and quality by assisting tasks in software product line (SPL). SPL decomposes a large-scale software system in terms of their functionalities. The goal of the decomposition is to create well-structured individual software systems that can meet different users’ requirements. Thus, feature modeling tools provides means to manage the inter-dependencies among reusable common and variable functionalities, called features. There are several tools to support variability management by modeling features in SPL. The variety of tools in the current literature makes it difficult to understand what kinds of tasks are supported and how much effort can be reduced by using these tools. In this paper, we present the results of an empirical study aiming to support SPL engineers choosing the feature modeling tool that best fits their needs. This empirical study compares and analyzes three tools, namely SPLOT, FeatureIDE , and pure::variants . These tools are analyzed based on data from 119 participants. Each participant used one tool for typical feature modeling tasks, such as create a model, update a model, automated analysis of the model, and product configuration. Finally, analysis concerning the perceived ease of use, usefulness, effectiveness, and efficiency are presented.

A Context-Aware Recommender System for Extended Software Product Line Configurations

Mass customization of standardized products has become a trend to succeed in todays market environment. Software Product Lines (SPLs) address this trend by describing a family of software products that share a common set of features. However, choosing the appropriate set of features that matches a users individual interests is hampered due to the overwhelming amount of possible SPL configurations. Recommender systems can address this challenge by filtering the number of configurations and suggesting a suitable set of features for the users requirements. In this paper, we propose a context-aware recommender system for predicting feature selections in an extended SPL configuration scenario, i.e. taking nonfunctional properties of features into consideration. We present an empirical evaluation based on a large real-world dataset of configurations derived from industrial experience in the Enterprise Resource Planning domain. Our results indicate significant improvements in the predictive accuracy of our context-aware recommendation approach over a state-of-the-art binary-based approach.

Visual guidance for product line configuration using recommendations and non-functional properties

Software Product Lines (SPLs) are a mature approach for the derivation of a family of products using systematic reuse. Different combinations of predefined features enable tailoring the product to fit the needs of each customer. These needs are related to functional properties of the system (optional features) as well as non-functional properties (e.g., performance or cost of the final product). In industrial scenarios, the configuration process of a final product is complex and the tool support is usually limited to check functional properties interdependencies. In addition, the importance of nonfunctional properties as relevant drivers during configuration has been overlooked. Thus, there is a lack of holistic paradigms integrating recommendation systems and visualizations that can help the decision makers. In this paper, we propose and evaluate an interrelated set of visualizations for the configuration process filling these gaps. We integrate them as part of the FeatureIDE tool and we evaluate its effectiveness, scalability, and performance.

N-dimensional tensor factorization for self-configuration of software product lines at runtime

Dynamic software product lines demand self-adaptation of their behavior to deal with runtime contextual changes in their environment and offer a personalized product to the user. However, taking user preferences and context into account impedes the manual configuration process, and thus, an efficient and automated procedure is required. To automate the configuration process, context-aware recommendation techniques have been acknowledged as an effective mean to provide suggestions to a user based on their recognized context. In this work, we propose a collaborative filtering method based on tensor factorization that allows an integration of contextual data by modeling an N-dimensional tensor User-Feature-Context instead of the traditional two-dimensional User-Feature matrix. In the proposed approach, different types of non-functional properties are considered as additional contextual dimensions. Moreover, we show how to self-configure software product lines by applying our N-dimensional tensor factorization recommendation approach. We evaluate our approach by means of an empirical study using two datasets of configurations derived for medium-sized product lines. Our results reveal significant improvements in the predictive accuracy of the configuration over a state-of-the-art non-contextual matrix factorization approach. Moreover, it can scale up to a 7-dimensional tensor containing hundred of configurations in a couple of milliseconds.

Personalized recommender systems for product-line configuration processes

Abstract Product lines are designed to support the reuse of features across multiple products. Features are product functional requirements that are important to stakeholders. In this context, feature models are used to establish a reuse platform and allow the configuration of multiple products through the interactive selection of a valid combination of features. Although there are many specialized configurator tools that aim to provide configuration support, they only assure that all dependencies from selected features are automatically satisfied. However, no support is provided to help decision makers focus on likely relevant configuration options. Consequently, since decision makers are often unsure about their needs, the configuration of large feature models becomes challenging. To improve the efficiency and quality of the product configuration process, we propose a new approach that provides users with a limited set of permitted, necessary and relevant choices. To this end, we adapt six state-of-the-art recommender algorithms to the product line configuration context. We empirically demonstrate the usability of the implemented algorithms in different domain scenarios, based on two real-world datasets of configurations. The results of our evaluation show that recommender algorithms, such as CF-shrinkage, CF-significance weighting, and BRISMF, when applied in the context of product-line configuration can efficiently support decision makers in a most efficient selection of features.