Computation with large-scale heterogeneous data typically requires universal traversal to search for all occurrences of a substructure that matches a possibly complex search pattern, whose context may be different in different places within the data. Both aspects cause difficulty for existing general-purpose programming languages, because these languages are designed for homogeneous data and have problems typing the different substructures in heterogeneous data, and the complex patterns to match with the substructures. Programmers either have to hard-code the structures and search patterns, preventing programs from being reusable and scalable, or have to use low-level untyped programming or programming with special-purpose query languages, opening the door to type mismatches that cause a high risk of program correctness and security problems. This thesis invents the concept of pattern structures, and proposes a general solution to the above problems - a programming technique using pattern structures. In this solution, well-typed pattern structures are defined to represent complex search patterns, and pattern searching over heterogeneous data is programmed with pattern parameters, in a statically-typed language that supports first-class typing of structures and patterns. The resulting programs are statically-typed, highly reusable for different data structures and different patterns, and highly scalable in terms of the complexity of data structures and patterns. Adding new kinds of patterns for an application no longer requires changing the language in use or creating new ones, but is only a programming task. The thesis demonstrates the application of this approach to, and its advantages in, two important examples of computation with heterogeneous data, i.e., XML data processing and Java bytecode analysis. / Thesis (Ph.D, Computing) -- Queen's University, 2007-08-27 09:43:38.888

This thesis is divided into a theoretical part, aimed at developing statements around the newly introduced concept of pattern morphisms, and a practical part, where we present use cases of pattern structures. A first insight of our work clarifies the facts on projections of pattern structures. We discovered that a projection of a pattern structure does not always lead again to a pattern structure. A solution to this problem, and one of the most important points of this thesis, is the introduction of pattern morphisms in Chapter4. Pattern morphisms make it possible to describe relationships between pattern structures, and thus enable a deeper understanding of pattern structures in general. They also provide the means to describe projections of pattern structures that lead to pattern structures again. In Chapter5 and Chapter6, we looked at the impact of morphisms between pattern structures on concept lattices and on their representations and thus clarified the theoretical background of existing research in this field. The application part reveals that random forests can be described through pattern structures, which constitutes another central achievement of our work. In order to demonstrate the practical relevance of our findings, we included a use case where this finding is used to build an algorithm that solves a real world classification problem of red wines. The prediction accuracy of the random forest is better, but the high interpretability makes our algorithm valuable. Another approach to the red wine classification problem is presented in Chapter 8, where, starting from an elementary pattern structure, we built a classification model that yielded good results.

La fusion d'informations consiste à résumer plusieurs informations provenant des différentes sources en une information exploitable et utile pour l'utilisateur.Le problème de la fusion est délicat surtout quand les informations délivrées sont incohérentes et hétérogènes. Les résultats de la fusion ne sont pas souvent exploitable et utilisables pour prendre une décision, quand ils sont imprécis. C'est généralement due au fait que les informations sont incohérentes. Plusieurs méthodes de fusion sont proposées pour combiner les informations imparfaites et elles appliquent l'opérateur de fusion sur l'ensemble de toutes les sources et considèrent le résultat tel qu'il est. Dans ce travail, nous proposons une méthode de fusion fondée sur l'Analyse Formelle de Concepts, en particulier son extension pour les données numériques : les structures de patrons. Cette méthode permet d'associer chaque sous-ensemble de sources avec son résultat de fusion. Toutefois l'opérateur de fusion est choisi, alors un treillis de concept est construit. Ce treillis fournit une classification intéressante des sources et leurs résultats de fusion. De plus, le treillis garde l'origine de l'information. Quand le résultat global de la fusion est imprécis, la méthode permet à l'utilisateur d'identifier les sous-ensemble maximaux de sources qui supportent une bonne décision. La méthode fournit une vue structurée de la fusion globale appliquée à l'ensemble de toutes les sources et des résultats partiels de la fusion marqués d'un sous-ensemble de sources. Dans ce travail, nous avons considéré les informations numériques représentées dans le cadre de la théorie des possibilités et nous avons utilisé trois sortes d'opérateurs pour construire le treillis de concepts. Une application dans le monde agricole, où la question de l'expert est d'estimer des valeurs des caractéristiques de pesticide provenant de plusieurs sources, pour calculer des indices environnementaux est détaillée pour évaluer la méthode de fusion proposée / Merging pieces of information into an interpretable and useful format is a tricky task even when an information fusion method is chosen. Fusion results may not be in suitable form for being used in decision analysis. This is generally due to the fact that information sources are heterogeneous and provide inconsistent information, which may lead to imprecise results. Several fusion operators have been proposed for combining uncertain information and they apply the fusion operator on the set of all sources and provide the resulting information. In this work, we studied and proposed a method to combine information using Formal Concept Analysis in particular Pattern Structures. This method allows us to associate any subset of sources with its information fusion result. Then once a fusion operator is chosen, a concept lattice is built. The concept lattice gives an interesting classification of fusion results and it keeps a track of the information origin. When the fusion global result is too imprecise, the method enables the users to identify what maximal subset of sources would support a more precise and useful result. Instead of providing a unique fusion result, the method yields a structured view of partial results labeled by subsets of sources. In this thesis, we studied the numerical information represented in the framework of possibility theory and we used three fusion operators to built the concept lattice. We applied this method in the context of agronomy when experts have to estimate several characteristics values coming from several sources for computing an environmental risk