Software processes relate to the sequence of steps that must be carried out by humans to pursue the goals of software engineering. In order to have an accurate representation of what these steps actually are, software processes can be modelled using a process modeling language (PML). Some PMLs simply support the specification of the steps, while others enable the process to be executed (or enacted). When enacted, software processes can provide guidance, automation and enforcement of the software engineering practices that are embodied in the model. Although there has been much fruitful research into PMLs, their adoption by industry has not been widespread. While the reasons for this lack of success may be many and varied, this thesis identified two areas in which PMLs may have been deficient: human dimension issues in terms of support for awareness and visualisation; and support for addressing management and resource issues that might arise dynamically when a process model is being enacted. In order to address some of these issues, a new visual PML called Virtual Reality Process Modelling Language (VRPML) has been developed and evaluated. Novel features have been introduced in VRPML to include support for the integration of a virtual environment, and dynamic creation and assignment of tasks and resources at the PML enactment level. VRPML serves as a research vehicle for addressing our main research hypothesis that a PML, which exploits a virtual environment, is useful to support software processes for distributed software engineering teams.

La validation temporelle des systèmes temps réel est nécessaire dans le cadre d’applications critiques tels l’aéronautique, le spatial ou l’automobile. Il s’agit, dans ces systèmes, de garantir les temps de réponse des tâches ainsi que le déterminisme de leurs communications. En raison de la complexité des systèmes actuels ainsi que de leur criticité, il est nécessaire de mettre en place une démarche de conception réduisant le temps de développement et ainsi le temps de mise en marché (time-to-market), tout en réduisant les risques d’erreurs de conception.Ce contexte rend l’ingénierie dirigée par les modèles particulièrement adaptée au développement de ce type de système. Les contributions de cette thèse partent des constats suivants. Premièrement, malgré la multitude des modèles d’analyses existants les modèles actuels ne peuvent retranscrire de nombreux cas rencontrés en industrie. Pour pouvoir analyser ces cas, il est nécessaire de les adapter `a l’analyse. Le deuxième constat porte sur l’adaptation qui n’est toujours pas simple surtout quand il existe une disparité sémantique entre les langages de description d’architecture et les modèles d’analyse ce qui nécessite de les rapprocher. Le dernier constat porte sur les difficultés dans la modélisation des systèmes distribués complexes car à moins de connaître le système complet, la représentation globale et sa validation temporelle reste coûteuse.Cette thèse propose des outils et méthodes pour améliorer le processus de modélisation et d’analyses temps réel. La première contribution consiste en la mise en place d’un référentiel de transformation de modèles endogène pour effectuer une adaptation conservative des modèles industriels aux modèles d’analyses. Dans l’optique de réduire l’écart sémantique entre les langages de description et les modèles d’analyse, cette thèse propose également une modélisation incrémentale des réseaux temps réel en vue de leur validation temporelle car les langages existants les considèrent de fa¸con limitée. La troisième contribution de la thèse porte sur la réduction des artéfacts de modélisation par extraction et élagage de méta-modèles afin d’obtenir les éléments nécessaires par rapport au points de vue d’analyse souhait´es.Toutes ces contributions sont implémentées dans des frameworks intégrant les processus d’analyses temps réel tels Time4Sys et MoSaRT et utilisées, dans le cadre d’un projet collaboratif, par des partenaires industriels. / The timing validation of real-time systems is mandatory for critical applications such as aeronautics,aerospace or automotive systems. The aim is to guarantee tasks response time and messages transmission time on networks. As for the criticality of these complex systems, it is necessary to implement a design process that reduces the development time therefore the time-to-market while reducing design errors risks.This context makes model-driven engineering well adapted for the development of critical real-time systems.The contributions of this thesis rely on the following observations. First of all, despite the existence of various analysis models, they often cannot represent perfectly some industrial cases. To analyze these cases,an adaptation is required to make them analyzable with existing tests. However, the adaptation is not quite easy especially in case of a semantic gap between systems description languages and analysis models. Also,several difficulties have been noticed to design and analyze an entire distributed complex system in one-shot unless knowing well the full system.In this PhD thesis, tools and methods are proposed to ease and improve the modeling and analysis processes of real-time systems. The first thesis contribution consists of implementing a rule-based endogenous transformation repository dedicated to adapt conservatively industrial models to the analysis models. The second contribution is focused on real-time networks and is dedicated to reduce the semantic gap between description languages and analysis models by proposing artefacts allowing to design networks on an incremental way. Moreover, this thesis proposes to reduce modeling artifacts using extraction and meta-models pruning inorder to retrieve useful elements referring to chosen analysis viewpoints.All these contributions are implemented in modeling frameworks integrating real-time analyses processes such as Time4Sys and MoSaRT, and used in a collaborative project by industrial partners.

Mémoire numérisé par la Division de la gestion de documents et des archives de l'Université de Montréal

No description available.

Process patterns allow the modular modelling and adaptable application of software processes. Present descriptions of process patterns show defects like non-uniform and unequivocal description forms and missing relationship definitions. These defects disadvantageously affect the effective usage of process patterns. In this work we present the language PROPEL (Process Pattern Description Language), which provides concepts for the semiformal description of process patterns and relationships between process patterns. With the help of PROPEL single process patterns can be modelled and, by definition of relationships, be composed to more complex processes. With the representation of different views of a process pattern catalog the process patterns and their relationships can be shown clearly. An example illustrates how a process pattern catalog and the contained process patterns are modelled. It is shown that in applying PROPEL the complexity of a process model can be reduced, the inconsistencies of processes can be eliminated and the flexibility of processes can be improved.

Domain-Specific Modelling Languages (DSMLs) offer constructs that are tailored to better capture the representational needs of an application domain. They provide customized graphical notations, which facilitate understanding of models by domain experts. As a result, DSMLs allow the construction of domain-specific models that support collaboration, improve work processes and enhance decision-making. Given their special purpose, however, a DSML has to be built each time a new application domain is to be addressed, which is quite demanding and time-consuming. A valid alternative is the creation of DSMLs through domain-specific adaptations of existing modelling languages. This solution has the benefits of starting from a baseline of well-known concepts, which is adapted to fit a specific purpose. Current engineering processes for building or adapting modelling languages, however, lack agility. It follows a sequential engineering lifecycle, where modelling and evaluation activities cannot start before the DSML is deployed for use. Such a sequential approach tends to keep the language engineer separate from the domain expert, who is hindered from gaining experience from the DSML until it is created. The separation of the two roles is a threat to the high quality of the DSML as it requires the joint effort of both experts. On the other hand, the new requirements that arise from the suggestions of the domain expert have to go through the whole engineering lifecycle (i.e. capture and document the requirement, conceptualise, implement and deploy), which is time-consuming. These current drawbacks of present engineering processes have been explored in two case studies, which report the development of a DSML for Patient Transferal Management and a DSML for Business Process as a Service. In this research an agile meta-modelling approach has been conceived to address the identified drawbacks. Specifically, the approach allows the quick interleaving of language engineering, modelling and evaluation activities. Hence, the close cooperation between the language engineers and the domain experts is fostered from an early stage. A set of operators are proposed to enable on-the-fly domain-specific adaptations of modelling languages, thus avoiding the sequential engineering phases. This agile meta-modelling aims to promote both the high-quality and quick development of DSMLs through domain-specific adaptations. Moreover, to avoid misinterpretation of the meaning of the newly created modelling constructs as well as ensuring machine interpretability of models, the agile meta-modelling has been supplemented by an ontology-aided approach. The latter embeds the specification specifications of modelling languages into an ontology. A set of semantic rules are proposed to support the propagation of language adaptations from the graphical to the machine-interpretable representation. In turn, the approach was developed in the modelling environment AOAME, which allows preserving consistency between the graphical and the machine-interpretable knowledge while domain-specific adaptations are performed. An evaluation strategy is proposed, from which three criteria were derived to evaluate the approach. Firstly, the correct design of the approach is evaluated by the extent to which it satisfies the requirements. Secondly, the operationability of the approach is evaluated by its ability to preserve consistency between the graphical and the machine-interpretable representations. Thirdly, the generality of the approach is evaluated by its ability to be applied in different application domains. The evaluation of operationability and generality are supported by implementing real-world use cases in AOAME. Consequently, the approach contributes to the practice in three different application domains, the Patient Transferal Management, Business Process as a Service and Innovation Processes. The scientific contribution of the approach spans research fields of Domain-Specific Modelling Language, Meta-Modelling, Enterprise Modelling and Ontologies. / Thesis (PhD)--University of Pretoria, 2019. / Informatics / PhD (Information Systems) / Unrestricted

With the popularity of model-driven methodologies and the abundance of modelling languages, a major question for a modeller is: Which language is suitable for modelling a system under study? To answer this question, one not only needs to know the range of relevant languages for modelling the system under study, but also needs to be able to compare these languages. In this dissertation, I consider these challenges from a semantic point of view for a diverse range of behavioural modelling languages that I refer to as the family of Big-Step Modelling Languages (BSMLs). There is a plethora of BSMLs, including statecharts, its variants, SCR, un-clocked variants of synchronous languages (e.g., Esterel and Argos), and reactive modules. BSMLs are often used to model systems that continuously interact with their environments. In a BSML model, the reaction of the model to an environmental input is a big step, which consists of a sequence of small steps, each of which can be the concurrent execution of a set of transitions. To provide a systematic method to understand and compare the semantics of BSMLs, this dissertation introduces the big-step semantic deconstruction framework that deconstructs the semantic design space of BSMLs into eight high-level, independent semantic aspects together with the enumeration of the common semantic options of each semantic aspect. The dissertation also presents a comparative analysis of the semantic options of each semantic aspect to assist one to choose one semantic option over another. A key idea in the big-step semantic deconstruction is that the high-level semantic aspects in the deconstruction recognize a big step as a whole, rather than only considering its constituent transitions operationally. A novelty of the big-step semantic deconstruction is that it lends itself to a systematic semantic formalization of most of the languages in the deconstruction. The dissertation presents a parametric, formal semantic definition method whose parameters correspond to the semantic aspects of the deconstruction, and thus it produces prescriptive semantics: The manifestation of a semantic option in the semantics of a BSML can be clearly identified. The way transitions are ordered to form a big step in a BSML is a source of semantic complexity: A modeller needs to be aware of the possible orders of the execution of transitions when constructing and analyzing a model. The dissertation introduces three semantic quality attributes that each exempts a modeller from considering an aspect of ordering in big steps. The ranges of BSMLs that support each of these semantic quality attributes are formally specified. These specifications indicate that achieving a semantic quality attribute in a BSML is a cross-cutting concern over the choices of its different semantic options. The semantic quality attributes together with the semantic analysis of individual semantic options can be used in tandem to assist a modeller or a semanticist to compare two BSMLs or to create a new, desired BSML from scratch. Through the big-step semantic deconstruction, I have discovered that some of the semantic aspects of BSMLs can be uniformly described as forms of synchronization. The dissertation presents a general synchronization framework for behavioural modelling languages. This framework is based on a notion of synchronization between transitions of complementary roles. It is parameterized by the number of interactions a transition can take part in, i.e., one vs. many, and the arity of the interaction mechanisms, i.e., exclusive vs. shared, which are considered for the complementary roles to result in 16 synchronization types. To enhance BSMLs with the capability to use the synchronization types, a synchronizer syntax is introduced for BSMLs, resulting in the family of Synchronizing Big-Step Modelling Languages (SBSMLs). Using the expressiveness of SBSMLs, the dissertation describes how underlying the semantics of many modelling constructs, such as multi-source, multi-destination transitions, various composition operators, and workflow patterns, there is a notion of synchronization that can be systematically modelled in SBSMLs.

With the popularity of model-driven methodologies and the abundance of modelling languages, a major question for a modeller is: Which language is suitable for modelling a system under study? To answer this question, one not only needs to know the range of relevant languages for modelling the system under study, but also needs to be able to compare these languages. In this dissertation, I consider these challenges from a semantic point of view for a diverse range of behavioural modelling languages that I refer to as the family of Big-Step Modelling Languages (BSMLs). There is a plethora of BSMLs, including statecharts, its variants, SCR, un-clocked variants of synchronous languages (e.g., Esterel and Argos), and reactive modules. BSMLs are often used to model systems that continuously interact with their environments. In a BSML model, the reaction of the model to an environmental input is a big step, which consists of a sequence of small steps, each of which can be the concurrent execution of a set of transitions. To provide a systematic method to understand and compare the semantics of BSMLs, this dissertation introduces the big-step semantic deconstruction framework that deconstructs the semantic design space of BSMLs into eight high-level, independent semantic aspects together with the enumeration of the common semantic options of each semantic aspect. The dissertation also presents a comparative analysis of the semantic options of each semantic aspect to assist one to choose one semantic option over another. A key idea in the big-step semantic deconstruction is that the high-level semantic aspects in the deconstruction recognize a big step as a whole, rather than only considering its constituent transitions operationally. A novelty of the big-step semantic deconstruction is that it lends itself to a systematic semantic formalization of most of the languages in the deconstruction. The dissertation presents a parametric, formal semantic definition method whose parameters correspond to the semantic aspects of the deconstruction, and thus it produces prescriptive semantics: The manifestation of a semantic option in the semantics of a BSML can be clearly identified. The way transitions are ordered to form a big step in a BSML is a source of semantic complexity: A modeller needs to be aware of the possible orders of the execution of transitions when constructing and analyzing a model. The dissertation introduces three semantic quality attributes that each exempts a modeller from considering an aspect of ordering in big steps. The ranges of BSMLs that support each of these semantic quality attributes are formally specified. These specifications indicate that achieving a semantic quality attribute in a BSML is a cross-cutting concern over the choices of its different semantic options. The semantic quality attributes together with the semantic analysis of individual semantic options can be used in tandem to assist a modeller or a semanticist to compare two BSMLs or to create a new, desired BSML from scratch. Through the big-step semantic deconstruction, I have discovered that some of the semantic aspects of BSMLs can be uniformly described as forms of synchronization. The dissertation presents a general synchronization framework for behavioural modelling languages. This framework is based on a notion of synchronization between transitions of complementary roles. It is parameterized by the number of interactions a transition can take part in, i.e., one vs. many, and the arity of the interaction mechanisms, i.e., exclusive vs. shared, which are considered for the complementary roles to result in 16 synchronization types. To enhance BSMLs with the capability to use the synchronization types, a synchronizer syntax is introduced for BSMLs, resulting in the family of Synchronizing Big-Step Modelling Languages (SBSMLs). Using the expressiveness of SBSMLs, the dissertation describes how underlying the semantics of many modelling constructs, such as multi-source, multi-destination transitions, various composition operators, and workflow patterns, there is a notion of synchronization that can be systematically modelled in SBSMLs.

Increasingly, client-server applications are implemented as web-based applications with user interfaces consisting entirely of web pages or equivalent renderings on other presentation channels (e.g. mobile or speech-based devices). However, the page-based medium and the stateless pull communication impose restrictions on the user interface that often manifest themselves in unsatisfactory dialog control, i.e. possibly severely diminished usability. We therefore present a Dialog Flow Notation that allows developers to encapsulate sequences of multiple dialog steps into reusable dialog modules that can be nested arbitrarily, and to specify different interaction patterns for different devices. The notation is complemented with a Dialog Control Framework that manages dialog flows on multiple channels, leaving only the tasks of implementing the device-independent application logic, designing the interface pages, and specifying the dialog flow to the developer.

This thesis presents the Multiple Modelling Quality Evaluation Framework method (hereinafter MMQEF), which is a conceptual, methodological, and technological framework for evaluating quality issues in modelling languages and modelling elements by the application of a taxonomic analysis. It derives some analytic procedures that support the detection of quality issues in model-driven projects, such as the suitability of modelling languages, traces between abstraction levels, specification for model transformations, and integration between modelling proposals. MMQEF also suggests metrics to perform analytic procedures based on the classification obtained for the modelling languages and artifacts under evaluation. MMQEF uses a taxonomy that is extracted from the Zachman framework for Information Systems (Zachman, 1987; Sowa and Zachman, 1992), which proposed a visual language to classify elements that are part of an Information System (IS). These elements can be from organizational to technical artifacts. The visual language contains a bi-dimensional matrix for classifying IS elements (generally expressed as models) and a set of seven rules to perform the classification. As an evaluation method, MMQEF defines activities in order to derive quality analytics based on the classification applied on modelling languages and elements. The Zachman framework was chosen because it was one of the first and most precise proposals for a reference architecture for IS, which is recognized by important standards such as the ISO 42010 (612, 2011). This thesis presents the conceptual foundation of the evaluation framework, which is based on the definition of quality for model-driven engineering (MDE). The methodological and technological support of MMQEF is also described. Finally, some validations for MMQEF are reported. / Esta tesis presenta el método MMQEF (Multiple Modelling Quality Evaluation Framework), el cual es un marco de trabajo conceptual, metodológico y tecnológico para evaluar aspectos de calidad sobre lenguajes y elementos de modelado mediante la aplicación de análisis taxonómico. El método deriva procedimientos analíticos que soportan la detección de aspectos de calidad en proyectos model-driven tales como: idoneidad de lenguajes de modelado, trazabilidad entre niveles de abstracción, especificación de transformación de modelos, e integración de propuestas de modelado. MMQEF también sugiere métricas para ejecutar procedimientos analíticos basados en la clasificación obtenida para los lenguajes y artefactos de modelado bajo evaluación. MMQEF usa una taxonomía para Sistemas de Información basada en el framework Zachman (Zachman, 1987; Sowa and Zachman, 1992). Dicha taxonomía propone un lenguaje visual para clasificar elementos que hacen parte de un Sistema de Información. Los elementos pueden ser artefactos asociados a niveles desde organizacionales hasta técnicos. El lenguaje visual contiene una matriz bidimensional para clasificar elementos de Sistemas de Información, y un conjunto de siete reglas para ejecutar la clasificación. Como método de evaluación MMEQF define actividades para derivar analíticas de calidad basadas en la clasificación aplicada sobre lenguajes y elementos de modelado. El marco Zachman fue seleccionado debido a que éste fue una de las primeras y más precisas propuestas de arquitectura de referencia para Sistemas de Información, siendo ésto reconocido por destacados estándares como ISO 42010 (612, 2011). Esta tesis presenta los fundamentos conceptuales del método de evaluación basado en el análisis de la definición de calidad en la ingeniería dirigida por modelos (MDE). Posteriormente se describe el soporte metodológico y tecnológico de MMQEF, y finalmente se reportan validaciones. / Aquesta tesi presenta el mètode MMQEF (Multiple Modelling Quality Evaluation Framework), el qual és un marc de treball conceptual, metodològic i tecnològic per avaluar aspectes de qualitat sobre llenguatges i elements de modelatge mitjançant l'aplicació d'anàlisi taxonòmic. El mètode deriva procediments analítics que suporten la detecció d'aspectes de qualitat en projectes model-driven com ara: idoneïtat de llenguatges de modelatge, traçabilitat entre nivells d'abstracció, especificació de transformació de models, i integració de propostes de modelatge. MMQEF també suggereix mètriques per executar procediments analítics basats en la classificació obtinguda pels llenguatges i artefactes de mode-lat avaluats. MMQEF fa servir una taxonomia per a Sistemes d'Informació basada en el framework Zachman (Zachman, 1987; Sowa and Zachman, 1992). Aquesta taxonomia proposa un llenguatge visual per classificar elements que fan part d'un Sistema d'Informació. Els elements poden ser artefactes associats a nivells des organitzacionals fins tècnics. El llenguatge visual conté una matriu bidimensional per classificar elements de Sistemes d'Informació, i un conjunt de set regles per executar la classificació. Com a mètode d'avaluació MMEQF defineix activitats per derivar analítiques de qualitat basades en la classificació aplicada sobre llenguatges i elements de modelatge. El marc Zachman va ser seleccionat a causa de que aquest va ser una de les primeres i més precises propostes d'arquitectura de referència per a Sistemes d'Informació, sent això reconegut per destacats estàndards com ISO 42010 (612, 2011). Aquesta tesi presenta els fonaments conceptuals del mètode d'avaluació basat en l'anàlisi de la definició de qualitat en l'enginyeria dirigida per models (MDE). Posteriorment es descriu el suport metodològic i tecnològic de MMQEF, i finalment es reporten validacions. / Giraldo Velásquez, FD. (2017). A framework for evaluating the quality of modelling languages in MDE environments [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/90628