Depuis les années 2000, l'informatique a vu émerger de nouvelles technologies, cloud et big data, qui bouleversent l'industrie avec l'arrivée d'outils de traitement à grande échelle.
De nouveaux besoins sont apparus comme la possibilité d'extraire de la valeur des données en s'appuyant sur des outils qui répondent aux nouvelles exigences technologiques.
Les architectures distribuées comme Hadoop, les bases de données non-relationnelles, les traitements parallélisés avec MapReduce constituent des outils qui répondent aux accroissements massifs des données, que ce soit en volumétrie, en nombre ou en type. Cette explosion de données a conduit à la terminologie Big Data.
Nous découvrirons les différents concepts des systèmes Big Data, ce que signifient les termes comme base NoSQL, MapReduce, lac de données, ETL ou ELT, etc.
Nous nous attarderons sur deux grands outils du BigData : Hadoop et MongoDB.[-]

Proposition d'une démarche pour le traitement de données complexes et/ou massives à des fins d'exploration interactive. Basée sur la mise en œuvre effective dans un contexte de données de santé, cette démarche propose d'explorer des notions connues mais peu utilisées qui émergent comme les bases graphes pour modéliser un lac de données et l'exploiter. Nous balayons quelques environnements système (Hadoop, bases NoSQL, ETL) et effleurons les contraintes de sécurité d'accès.[-]

After giving a motivation of graph databases and an overview of the main data models, we will dive into foundational aspects of graph database query languages, with a strong focus on regular path queries (RPQs) and conjunctive regular path queries (CRPQs). We will consider the different semantics that graph database systems use for such queries (every path, simple path, trail), and we will look into the computational complexities of query evaluation and query containment.
After having gone through these foundations, we plan to do some excursions into connections between tree-structured and graph-structured data, adding data value comparisons, and aspects of real-life queries.[-]

Semistructured data is an umbrella term encompassing data models which are not logically organized in tables (i.e., the relational data model) but rather in hierarchical structures using markers such as tags to separate semantic elements and data fields in a ‘self-describing' way. In this lecture we survey some of the multiple connections between formal language theory and semi-structured data, in particular concerning the XML format. We will cover ranked and unranked tree automata, and its connections to Monadic Second Order logic, First Order logic, and XPath. The aim is to take a glimpse at the landscape of closure properties, algorithms and expressiveness results for these formalisms.[-]

After giving a motivation of graph databases and an overview of the main data models, we will dive into foundational aspects of graph database query languages, with a strong focus on regular path queries (RPQs) and conjunctive regular path queries (CRPQs). We will consider the different semantics that graph database systems use for such queries (every path, simple path, trail), and we will look into the computational complexities of query evaluation and query containment.
After having gone through these foundations, we plan to do some excursions into connections between tree-structured and graph-structured data, adding data value comparisons, and aspects of real-life queries.[-]

Semistructured data is an umbrella term encompassing data models which are not logically organized in tables (i.e., the relational data model) but rather in hierarchical structures using markers such as tags to separate semantic elements and data fields in a ‘self-describing' way. In this lecture we survey some of the multiple connections between formal language theory and semi-structured data, in particular concerning the XML format. We will cover ranked and unranked tree automata, and its connections to Monadic Second Order logic, First Order logic, and XPath. The aim is to take a glimpse at the landscape of closure properties, algorithms and expressiveness results for these formalisms.[-]