In Software Product Lines (SPLs), a feature model can be used to represent the similarities and differences within a family of software systems. This allows describing the systems derived from the product line as a unique combination of the features in the model. What makes feature models particularly appealing is the fact that the constraints in the model prevent incompatible features from being part of the same product. Despite the benefits of feature models, constructing and maintaining these models can be a laborious task especially in product lines with a large number of features and constraints. As a result, the study of automated techniques to reason on feature models has become an important research topic in the SPL community in recent years. Two techniques, in particular, have significant appeal for researchers: SAT solvers and Binary Decision Diagrams (BDDs). Each technique has been applied successfully for over four decades now to tackle many practical combinatorial problems in various domains. Currently, several approaches have proposed the compilation of feature models to specific logic representations to enable the use of SAT solvers and BDDs. In this thesis, we argue that several critical issues related to the use of SAT solvers and BDDs have been consistently neglected. For instance, satisfiability is a well-known NP-complete problem which means that, in theory, a SAT solver might be unable to check the satisfiability of a feature model in a feasible amount of time. Similarly, it is widely known that the size of BDDs can become intractable for large models. At the same time, we currently do not know precisely whether these are real issues when feature models, especially large ones, are compiled to SAT and BDD representations. Therefore, in our research we provide a significant step forward in the state-of-the-art by examining deeply many relevant properties of the feature modeling domain and the mechanics of SAT solvers and BDDs and the sensitive issues related to these techniques when applied in that domain. Specifically, we provide more accurate explanations for the space and/or time (in)tractability of these techniques in the feature modeling domain, and enhance the algorithmic performance of these techniques for reasoning on feature models. The contributions of our work include the proposal of novel heuristics to reduce the size of BDDs compiled from feature models, several insights on the construction of efficient domain-specific reasoning algorithms for feature models, and empirical studies to evaluate the efficiency of SAT solvers in handling very large feature models.

Made available in DSpace on 2014-06-12T15:54:05Z (GMT). No. of bitstreams: 2 arquivo6551_1.pdf: 2254714 bytes, checksum: 89a6702d1c801f178299f95585aac5ab (MD5) license.txt: 1748 bytes, checksum: 8a4605be74aa9ea9d79846c1fba20a33 (MD5) Previous issue date: 2007 / Linha de Produtos de Software (LPS) é uma aborgadem promissora para o desenvolvimento de um conjunto de produtos focados em um segmento de mercado e desenvolvidos a partir de um conjunto comum de artefatos. Possíveis benefícios incluem reuso em larga escala e significativa melhoria em produtividade. Um problema-chave associado, no entanto, é o tratamento de estratégias de implantação, em que uma organização decide iniciar uma LPS a partir do zero, fazer bootstrap de produtos existentes em uma LPS, ou evoluir uma LPS. Em particular, no nível de implementação e de modelo de features, métodos de desenvolvimento carecem de apoio adequado para extração e evolução de LPSs. Neste contexto, apresentamos um m´etodo original provendo diretrizes concretas para extração e evolução de LPSs no nível de implementação e de modelo de features, nos quais proporciona reuso e segurança. O método primeiro faz o bootstrap da LPS e então a evolui com uma abordagem reativa. O método se baseia em uma coleção de refatoramentos tanto na implementação (refatoramentos orientados a aspectos) como no modelo de features. O método foi avaliado no domínio altamente variável de jogos móveis

Une particularité importante du logiciel est sa capacité à être adapté et configuré selon différents scénarios. Récemment, la variabilité du logiciel a été étudiée comme un concept de première classe dans différents domaines allant des lignes de produits logiciels aux systèmes ubiquitaires. La variabilité est la capacité d'un produit logiciel à varier en fonction de différentes circonstances. Les systèmes à forte variabilité mettent en jeu des produits logiciels où la gestion de la variabilité est une activité d'ingénierie prédominante. Les diverses parties de ces systèmes sont couramment modélisées en utilisant des formes différentes de ''modèle de variabilité'', qui est un formalisme de modélisation couramment utilisé. Les modèles de caractéristiques (feature models) ont été introduits par Kang et al. en 1990 et sont une représentation compacte d'un ensemble de configurations pour un système à forte variabilité. Le grand nombre de configurations d'un modèle de caractéristiques ne permet pas une analyse manuelle. De fait, les mécanismes assistés par ordinateur sont apparus comme une solution pour extraire des informations utiles à partir de modèles de caractéristiques. Ce processus d'extraction d'information à partir de modèles de caractéristiques est appelé dans la littérature scientifique ''analyse automatisée de modèles de caractéristiques'' et a été l'un des principaux domaines de recherche ces dernières années. Plus de trente opérations d'analyse ont été proposées durant cette période. Dans cette thèse, nous avons identifié différentes questions ouvertes dans le domaine de l'analyse automatisée et nous avons considéré plusieurs axes de recherche. Poussés par des scénarios du monde réel (e.g., la téléphonie mobile ou la vidéo protection), nous avons contribué à appliquer, adapter ou étendre des opérations d'analyse automatisée pour l’évolution, le test et la configuration de systèmes à forte variabilité. / The large number of configurations that a feature model can encode makes the manual analysis of feature models an error prone and costly task. Then, computer-aided mechanisms appeared as a solution to extract useful information from feature models. This process of extracting information from feature models is known as ''Automated Analysis of Feature models'' that has been one of the main areas of research in the last years where more than thirty analysis operations have been proposed. In this dissertation we looked for different tendencies in the automated analysis field and found several research opportunities. Driven by real-world scenarios such as smart phone or videosurveillance domains, we contributed applying, adapting or extending automated analysis operations in variability intensive systems evolution, testing and configuration.

No description available.

abstract: Speech is known to serve as an early indicator of neurological decline, particularly in motor diseases. There is significant interest in developing automated, objective signal analytics that detect clinically-relevant changes and in evaluating these algorithms against the existing gold-standard: perceptual evaluation by trained speech and language pathologists. Hypernasality, the result of poor control of the velopharyngeal flap---the soft palate regulating airflow between the oral and nasal cavities---is one such speech symptom of interest, as precise velopharyngeal control is difficult to achieve under neuromuscular disorders. However, a host of co-modulating variables give hypernasal speech a complex and highly variable acoustic signature, making it difficult for skilled clinicians to assess and for automated systems to evaluate. Previous work in rating hypernasality from speech relies on either engineered features based on statistical signal processing or machine learning models trained end-to-end on clinical ratings of disordered speech examples. Engineered features often fail to capture the complex acoustic patterns associated with hypernasality, while end-to-end methods tend to overfit to the small datasets on which they are trained. In this thesis, I present a set of acoustic features, models, and strategies for characterizing hypernasality in dysarthric speech that split the difference between these two approaches, with the aim of capturing the complex perceptual character of hypernasality without overfitting to the small datasets available. The features are based on acoustic models trained on a large corpus of healthy speech, integrating expert knowledge to capture known perceptual characteristics of hypernasal speech. They are then used in relatively simple linear models to predict clinician hypernasality scores. These simple models are robust, generalizing across diseases and outperforming comprehensive set of baselines in accuracy and correlation. This novel approach represents a new state-of-the-art in objective hypernasality assessment. / Dissertation/Thesis / Masters Thesis Electrical Engineering 2020

Mobile applications are developed to run on fast-evolving platforms, such as Android or iOS. Respective mobile devices are heterogeneous concerning hardware (e.g., sensors, displays, communication interfaces) and software, especially operating system functions. Software vendors cope with platform evolution and various hardware configurations by abstracting from these variable assets. However, they cannot be sure about their assumptions on the inner conformance of all device parts and that the application runs reliably on each of them—in consequence, comprehensive testing is required. Thereby, in testing, variability becomes tedious due to the large number of test cases required to validate behavior on all possible device configurations. In this paper, we provide remedy to this problem by combining model-based testing with variability concepts from Software Product Line engineering. For this purpose, we use feature-based test modeling to generate test cases from variable operational models for individual application configurations and versions. Furthermore, we illustrate our concepts using the commercial mobile application “runtastic” as example application.

Domain feature models concisely express commonality and variability among variants of a software product line. For separation of concerns, e.g., due to legal restrictions, technical considerations, and business requirements, multi-view approaches restrict the configuration choices on feature models for different stakeholders. However, recent approaches lack a formalization for precise, yet flexible specifications of views that ensure every derivable configuration perspective to obey feature model semantics. Here, we introduce a novel approach for clustering feature models to create multi-perspectives. Such customized perspectives result from composition of multiple concern-relevant views. A structured view model is used to organize feature groups, whereat a feature can be contained in multiple views. We provide formalizations for view composition and guaranteed consistency of the resulting perspectives w.r.t. feature model semantics. Thereupon, an efficient algorithm to verify consistency for entire clusterings is provided. We present an implementation and evaluate our concepts on two case studies.

Les Matrices de Comparaison de Produits (MCP) sont largement répandues sur le web. Elles fournissent une représentation simple des caractéristiques d'un ensemble de produits. Cependant, le manque de formalisation et la grande diversité des MCP rendent difficile le développement de logiciels pour traiter ces matrices. Dans cette thèse, nous développons deux approches complémentaires pour la formalisation de MCP. La première consiste en une description précise de la structure et la sémantique d'une MCP sous la forme d'un meta-modèle. Nous proposons aussi une transformation automatique d'une MCP vers un modèle de MCP conforme au meta-modèle. La seconde approche consiste à synthétiser des modèles de caractéristiques attribués à partir d'une classe de MCP. Grâce nos contributions, nous proposons une approche générique et extensible pour la formalisation et l'exploitation de MCP. / Product Comparison Matrices (PCMs) abound on the Web. They provide a simple representation of the characteristics of a set of products. However, the lack of formalization and the large diversity of PCMs challenges the development of software for processing these matrices. In this thesis, we develop two complementary approaches for the formalisation of PCMs. The first one consists in a precise description of the structure and semantics of PCMs in the form of a metamodel. We also propose an automated transformation from PCMs to PCM models conformant to the metamodel. The second one consists in synthesizing attributed feature models from a class of PCMs. With our contributions, we propose a generic and extensible approach for the formalization and exploitation of PCMs.

Mobile applications are developed to run on fast-evolving platforms, such as Android or iOS. Respective mobile devices are heterogeneous concerning hardware (e.g., sensors, displays, communication interfaces) and software, especially operating system functions. Software vendors cope with platform evolution and various hardware configurations by abstracting from these variable assets. However, they cannot be sure about their assumptions on the inner conformance of all device parts and that the application runs reliably on each of them—in consequence, comprehensive testing is required. Thereby, in testing, variability becomes tedious due to the large number of test cases required to validate behavior on all possible device configurations. In this paper, we provide remedy to this problem by combining model-based testing with variability concepts from Software Product Line engineering. For this purpose, we use feature-based test modeling to generate test cases from variable operational models for individual application configurations and versions. Furthermore, we illustrate our concepts using the commercial mobile application “runtastic” as example application.

Modern business applications are getting increasingly distributed as multi-tenant software as a service (SaaS). This leads to new challenges in terms of quality assurance, because all customers are directly affected by software changes. The resulting problem is to proactively determinate evolutionary effects. Because SaaS applications are often realized in the sense of a software product line, this thesis examines ways of using feature models to face the mentioned problem. For this purpose, two approaches are analyzed: extended feature models with quality attributes annotated per feature and the analysis of structural aspects of feature models and corresponding concrete configurations. The presented attributed feature model approach measures the quality of concrete configurations to make configurations comparable according to specific quality goals. Criteria are elicited for when configurations can be compared to draw helpful conclusions. The structural approach focuses economic questions that are quality assurance related, such as identifying features that none of the tenants selected in their application configurations. Furthermore, three algorithms are presented that demonstrate the structural analysis approach to gather information relevant to quality assurance.