Sonia Guehis

Data-Driven Publication of Relational Databases

The paper presents a framework for publishing relational databases in textual documents such as mails, HTML pages, LATEX or BibTex files, plain texts, etc. The publication process relies on a mapping of the relational database to a virtual data graph which supports navigation operators. Applications can express the data they need by navigating in the graph. These operations are provided by a declarative query language over virtual graphs, named DOCQL. We illustrate its features with the conference management system MYREVIEW

Publish By Example

We propose an approach for producing database publishing programs by example. The main idea is to interactively build an example document, representative of the program output. The system infers from this document, without ambiguity, the publishing program. The end-user does not need to know a programming language, a query language or the database schema.

Big Data Analytic Approaches Classification.

Analytical data management applications, affected by the explosion of the amount of generated data in the context of Big Data, are shifting away their analytical databases towards a vast landscape of architectural solutions combining storage techniques, programming models, languages, and tools. To support users in the hard task of deciding which Big Data solution is the most appropriate according to their specific requirements, we propose a generic architecture to classify analytical approaches. We also establish a classification of the existing query languages, based on the facilities provided to access the big data architectures. Moreover, to evaluate different solutions, we propose a set of criteria of comparison, such as OLAP support, scalability, and fault tolerance support. We classify different existing Big Data analytics solutions according to our proposed generic architecture and qualitatively evaluate them in terms of the criteria of comparison. We illustrate how ou r proposed generic architecture can be used to decide which Big Data analytic approach is suitable in the context of several use cases.

Speeding-up data-driven applications with program summaries

We consider the class of database programs and address the problem of minimizing the cost of their exchanges with the database server. This cost partly consists of query execution at the server side, and partly of query submission and network exchanges between the program and the server. The natural organization of database programs leads to submit an intensive flow of elementary SQL queries to the server, and exploits only locally its optimization power. In this paper, we develop a global optimization approach. We base this approach on an execution model where queries can be executed asynchronously with respect to the flow of the application program. Our method aims at choosing an efficient query scheduling which limits the penalty of client/server interactions. Our results show that the technique can improve the execution time of database programs by several orders of magnitude.

Un modèle de production interactive de programmes de publication

We propose an approach for producing database publishing programs by example. The main idea is to interactively build a canonical document, representative of the program output. The system infers from this document, without ambiguity, the publishing program. The end-user does not need to know a programming language, a query language or the database schema. We describe and comment a visual editor that shows how a user can rely on these concepts for intuitively producing publishing programs.

Classifying Big Data Analytic Approaches: A Generic Architecture

The explosion of the huge amount of generated data to be analyzed by several applications, imposes the trend of the moment, the Big Data boom, which in turn causes the existence of a vast landscape of architectural solutions. Non expert users who have to decide which analytical solutions are the most appropriates for their particular constraints and specific requirements in a Big Data context, are today lost, faced with a panoply of disparate and diverse solutions. To support users in this hard selection task, in a previous work, we proposed a generic architecture to classify Big Data Analytical Approaches and a set of criteria of comparison/evaluation. In this paper, we extend our classification architecture to consider more types of Big Data analytic tools and approaches and improve the list of criteria to evaluate them. We classify different existing Big Data analytics solutions according to our proposed generic architecture and qualitatively evaluate them in terms of the criteria of comparison. Additionally, we propose a preliminary design of a decision support system, intended to generate suggestions to users based on such classification and on a qualitative evaluation in terms of previous users experiences, users requirements, nature of the analysis they need, and the set of evaluation criteria.