Geneviève Pujolle

Multidimensional database design from document-centric XML documents

Despite a decade of research in OLAP systems, very few works attempt to tackle the problem of analysing data extracted from XML text-rich documents. These documents are loosely structured XML documents mainly composed of text. This paper details conceptual design steps of multidimensional databases from such documents. With the use of an adapted multidimensional conceptual model, the design process allows the integration of data extracted from text-rich XML documents within an adapted OLAP system.

Recommandation de requêtes dans les bases de données multidimensionnelles annotées

This paper deals with personalization of annotated OLAP systems. Constellation data are extended with annotations and user preferences. Annotations reflect the decision-maker experience, whereas preferences enable users to focus on the most interesting data. User preferences allow annotated contextual recommendations that assist the decision-maker during hislher multidimensional navigations.

Fonctions d'agrégation pour l'analyse en ligne (OLAP) de données textuelles. Fonctions TOP_KWk et AVG_KW opérant sur des termes.

For more than a decade, researches on OLAP and Multidimensional DataBases (MDB) have generated methodologies, tools and resource management systems for the analysis of numerical data. With the growing availability of digital documents, there is a need for incorporating text-rich documents within multidimensional databases as well as an adapted framework for their analysis. This paper presents new aggregation functions that allow aggregating textual data within the OLAP environment. On the one hand, the TOP_KEYWORD function (TOP_KW for short) summarises a set of documents by their most significant terms. On the other hand, the AVERAGE_KEYWORD (AVG_KW for short) attempts to summarise sets of keywords by more general terms while controlling the loss of sense implied by this aggregation.