Volume 13, Number 5/6

Considering Structural and Vocabulary Heterogeneity in XML Query: FPTPQ and Holistic Evaluation.

  Authors

Brice Nguefack1, Maurice Tchoupé Tchendji1 and Thomas Djotio Ndie2, 1University of Dschang, Cameroon, 2University of Yaoundé I, Cameroun

  Abstract

The integration of XML data sources which have different schemas/DTD can originate structural and vocabular heterogeneity. In this context, it is difficult to write satisfiable queries. As a solution, many Information Systems focus on building approximate evaluation techniques for exact queries. As a project, we build flexible and preference XML query languages and associated evaluation algorithms. In this paper, we propose the Flexible Preference Tree Pattern Query (FPTPQ), a new TPQ that allows multiple items/names (resp. paths) for the same node, in order to integrate (resp. to locate) all the different instances of the database nodes. The FPTPQ enable to have preference nodes and ordering operators among label items and paths. We also provide a holistic algorithm that evaluates the FPTPQ and capitalises the preferences to determine the best available solutions. Illustrations and experimentations are realized to show the effectiveness of our solutions.

  Keywords

XML Query with structural preference, Structural heterogeneity, vocabular heterogeneity, flexible query, holistic matching algorithm, ranked results.