Concurrency is generally considered to be difficult due to a lack of appropriate abstraction, rather than inherent complexity. Lock-based approaches to mutual exclusion are pervasive, despite the presence of models that are easier to understand, such as the message-passing model present in CSP (Communicating Sequential Processes). CSP provides a rich framework for building and reasoning about concurrent systems, but has historically required a change of programming language or paradigm in order to work with it. The Go programming language is a modern, imperative programming language that includes native support for processes and channels. The popularity of this language has grown and more and more people are being exposed to the fundamental ideas of CSP. There is a gap in the understanding of how a restrictive formal model can interact with and support the development of concurrent programs in a language such as Go. Through a series of case studies and analysis, we show how the CSP concurrency model can be used as the basis for the design of a concurrent system architecture without requiring the program to be written entirely as the composition of processes. It is also possible to use the CSP process algebra to build abstract models and use model-checking tools to verify properties of a concurrent system. These models can then be used to guide the decomposition of a system into a more fine-grained concurrent system. This thesis bridges the gap between the development of CSP-style concurrent software and the formal model of CSP. In particular, it shows how it is not necessary for a program or programming language to conform to rigid structure in order for CSP to be a useful tool for the development of reliable and easy to understand concurrent systems.

Cette thèse a pour but d’étudier des propriétés structurelles d’automates étendant celle du déterminisme, et les langages pouvant être dénotés par une expression rationnelle dont l’automate des positions présente l’une de ces propriétés. Si Book et al. ont montré que tous les langages rationnels peuvent être reconnus par un automate des positions non-ambigu, Brüggemann-Klein et Wood ont montré que ceux pouvant l’être par un automate des positions déterministe forment une famille strictement incluse dans celle des rationnels. Nous nous intéressons aux extensions de cette famille, en cherchant à caractériser leurs langages, et à étudier leur hiérarchie interne et leur inclusion entre elles. / This thesis aims to study structural properties of automata extending determinism, and the languages that can be denoted by a regular expression of which the position automaton has one such property. If Book et al. showed that all regular languages can be recognized by an unambiguous position automaton, Brüggemann-Klein and Wood showed that only a proper subset of them can be recognized by a deterministic position automaton. We focus on extensions of this subfamily, by seeking to characterize their languages, and to study their internal hierarchy and how they relate to each other.

This thesis describes the Omnibus language and its supporting framework of tools. Omnibus is an object-oriented language which is superficially similar to the Java programming language but uses value semantics for objects and incorporates a behavioural interface specification language. Specifications are defined in terms of a subset of the query functions of the classes for which a frame-condition logic is provided. The language is well suited to the specification of modelling types and can also be used to write implementations. An overview of the language is presented and then specific aspects such as subtleties in the frame-condition logic, the implementation of value semantics and the role of equality are discussed. The challenges of reference semantics are also discussed. The Omnibus language is supported by an integrated verification tool which provides support for three assertion-based verification approaches: run-time assertion checking, extended static checking and full formal verification. The different approaches provide different balances between rigour and ease of use. The Omnibus tool allows these approaches to be used together in different parts of the same project. Guidelines are presented in order to help users avoid conflicts when using the approaches together. The use of the integrated verification approach to meet two key requirements of safe software component reuse, to have clear descriptions and some form of certification, are discussed along with the specialised facilities provided by the Omnibus tool to manage the distribution of components. The principles of the implementation of the tool are described, focussing on the integrated static verifier module that supports both extended static checking and full formal verification through the use of an intermediate logic. The different verification approaches are used to detect and correct a range of errors in a case study carried out using the Omnibus language. The case study is of a library system where copies of books, CDs and DVDs are loaned out to members. The implementation consists of 2278 lines of Omnibus code spread over 15 classes. To allow direct comparison of the different assertion-based verification approaches considered, run-time assertion checking, extended static checking and then full formal verification are applied to the application in its entirety. This directly illustrates the different balances between error coverage and ease-of-use which the approaches offer. Finally, the verification policy system is used to allow the approaches to be used together to verify different parts of the application.

Vérifier formellement des programmes concurrents est une tâche difficile. S’il existe différentes techniques pour la réaliser, très peu sont effectivement mises en oeuvre pour des programmes écrits dans des langages de programmation réalistes. En revanche, les techniques de vérification formelles de programmes séquentiels sont utilisées avec succès depuis plusieurs années déjà, et permettent d’atteindre de hauts degrés de confiance dans nos systèmes. Cette thèse propose une alternative aux méthodes d’analyses dédiées à la vérification de programmes concurrents consistant à transformer le programme concurrent en un programme séquentiel pour le rendre analysable par des outils dédiés aux programmes séquentiels. Nous nous plaçons dans le contexte de FRAMA-C, une plate-forme d’analyse de code C spécifié avec le langage ACSL. Les différentes analyses de FRAMA-C sont des greffons à la plate-forme, ceux-ci sont à ce jour majoritairement dédiés aux programmes séquentiels. La méthode de vérification que nous proposons est appliquée manuellement à la vérification d’un code concurrent issu d’un hyperviseur. Nous automatisons la méthode à travers un nouveau greffon à FRAMA-C qui permet de produire automatiquement, depuis un programme concurrent spécifié, un programme séquentiel spécifié équivalent. Nous présentons les bases de sa formalisation, ayant pour but d’en prouver la validité. Cette validité n’est valable que pour la classe des programmes séquentiellement consistant. Nous proposons donc finalement un prototype de solveur de contraintes pour les modèles mémoire faibles, capable de déterminer si un programme appartient bien à cette classe en fonction du modèle mémoire cible. / Formal verification of concurrent programs is a hard task. There exists different methods to perform such a task, but very few are applied to the verification of programs written using real life programming languages. On the other side, formal verification of sequential programs is successfully applied for many years, and allows to get high confidence in our systems. As an alternative to dedicated concurrent program analyses, we propose a method to transform concurrent programs into sequential ones to make them analyzable by tools dedicated to sequential programs. This work takes place within the analysis framework FRAMA-C, dedicated to the analysis of C code specified with ACSL. The different analyses provided by FRAMA-C are plugins to the framework, which are currently mostly dedicated to sequential programs. We apply this method to the verification of a concurrent code taken from an hypervisor. We describe the automation of the method implemented by a new plugin to FRAMAC that allow to produce, from a specified concurrent program, an equivalent specified sequential program. We present the basis of a formalization of the method with the objective to prove its validity. This validity is admissible only for the class of sequentially consistent programs. So, we finally propose a prototype of constraint solver for weak memory models, which is able to determine whether a program is in this class or not, depending on the targeted hardware.

Cette thèse propose une méthode pour alimenter une ontologie, une structure de concepts liés par des relations sémantiques, par des termes français, anglais et allemands à partir de corpus spécialisés comparables. Son apport principal est le développement des méthodes d'extraction utilisant des ressources endogènes apprises à partir de corpus et d'ontologie. Exploitant des n-grammes de caractères, elles sont disponibles et indépendantes vis-à-vis de la langue et du domaine. La première contribution porte sur l'utilisation des ressources morphologiques et morphosyntaxiques endogènes pour extraire des termes mono- et polylexicaux à partir de corpus. La deuxième contribution vise à exploiter des ressources endogènes pour identifier leurs traductions. La troisième contribution concerne la construction des familles morphologiques endogènes servant à alimenter l'ontologie. / This thesis proposes a method of enrichment and population of an ontology, a structure of concepts linked by semantic relations, by terms in French, English and German from comparable domain-specific corpora. Our main contribution is the development of extraction methods based on endogenous resources, learned from the corpus and the ontology being analyzed. Using caracter n-grams, these resources are available and independent of a particular language or domain. The first contribution concerns the use of endogenous morphological and morphosyntactic resources for mono- and polylexical terms extraction from the corpus. The second contribution aims to use endogenous resources to identify translations for these terms. The third contribution concerns the construction of endogenous morphological families designed to enrich and populate the ontology.

Dans cette thèse, nous nous intéressons aux aspects algorithmiques et applications de deux problèmes de graphes, à savoir, la coloration et les alliances. La première partie concerne deux variantes de la coloration de graphes, la coloration Grundy et la coloration forte stricte. Nous commençons par l'étude du nombre Grundy des graphes réguliers. Nous donnons une condition fixe k, nous fournissons une condition nécessaire et suffisante pour que le nombre Grundy d'un graphe régulier soit au moins égal k. Nous caractérisons la classe des graphes cubiques (3-réguliers) pour laquelle le nombre Grundy est égal à 4, et nous présentons un algorithme linéaire pour déterminer le nombre Grundy d'un graphe cubique quelconque. Par ailleurs, en se basant sur la coloration forte stricte pour décomposer les arbres en petites composantes, nous présentons un nouvel algorithme pour l'appariement d'arbres étiquetés, non-ordonnés non-enracinés. Nous montrons que la distance calculée entre deux arbres est une pseudo-métrique. Nos expérimentations sur de larges bases synthétiques et des bases de données réelles confirment nos résultats analytiques et montrent que la distance proposée est précise et son algorithme est scalable. La seconde partie de cette thèse est consacrée aux alliances dans les graphes. Nous proposons un algorithme distribué autostabilisant pour la construction d'alliance offensive globale minimale dans un graphe arbitraire. Nous démontrons que cet algorithme converge sous le démon synchrone en temps linéaire. Ensuite, nous donnons le premier algorithme distribué autostabilisant pour le problème de l'alliance forte globale minimale dans un graphe quelconque. Nous prouvons que cet algorithme est polynomial sous le démon inéquitable distribué. Nous montrons par la suite, comment cet algorithme peut être adapté pour des généralisations du problème, comme la k-alliance forte et l'alliance forte pondérée. Enfin, en se basant sur les propriétés structurelles de l'alliance offensive, nous présentons une solution pour décentraliser le protocole de signalisation SIP. Ceci rend possible son déploiement dans un réseau mobile ad hoc / This thesis investigates the algorithmic aspects and applications of two graph problems, namely, colorings and alliances. In the first part, we focus on two variants of the proper vertex coloring, the Grundy coloring and the strict strong coloring. We start by the study of Grundy number for regular graphs. We give a sufficient condition for d-regular graphs with sufficiently large girth to have Grundy number equals d + 1. Then, using graph homomorphism, we obtain a necessary and sufficient condition for d-regular graphs to have Grundy number at least k. Moreover, we characterize cubic graphs (3-regular) for which the Grundy number is d + 1, and present a linear-time algorithm to determine the Grundy number of any arbitrary cubic graph. Subsequently, based on the strict strong coloring, we present an approach for the problem of matching labeled trees. Using this coloring, we propose a new algorithm to deterministically decompose a tree into small components. This leads to an efficient algorithm to measure an accurate distance between unrooted unordered labeled trees. The second part is devoted to the alliances in graphs. We first propose a linear-time self-stabilizing algorithm for the minimal global offensive alliance set problem, under the synchronous distributed scheduler. Then, we give the first self-stabilizing algorithm for the minimal global powerful alliance set problem in arbitrary graphs. Moreover, we show how this algorithm can be adapted to find the minimal global powerful k-alliance and the minimal weighted global powerful alliance sets. We prove that all these algorithms converge in polynomial-time under the unfair distributed scheduler. Finally, based on the structural properties of the offensive alliance, we propose a solution to decentralize the signaling protocol SIP. This enables SIP applications in mobile ad hoc networks