Research Lines

Document management, storage and retrieval

This research concerns the modelling, organisation and storage of structured documents. We mainly focus on representing and querying the temporal dimensions of documents, such as their publication dates and event-time periods.

Knowledge(-based) extraction

The purpose of this reseach line is to desing fast algorithms for the extraction of useful data and knowledge from unstructured texts. Our approach uses graph-matching and ontology inference to recognise the ontology concepts and instances, avoiding the use of Natural Language Processing.

Ontology Learning

In this research line we analyse the application of document clustering and data mining algorithms in order to automatically extend and populate domain ontologies.

Multidimensional analysis of semi-structured documents

The aim of this research line is to define a series of tools for the design, creation, and loading of semi-structured multidimensional models to allow the analysis of web resource sets with high textual contents (XML, RDF u OWL), as well as their mining for the discovering of new knowledge that can be helpful for virtual organizations in their decision making.

Semantic GRID

In this research line we investigate the application of knowledge management tools and database techniques to the development, discovery, analysis and organization of GRID services according to the semantics of its application domain.

Temporal Information Analysis

This research comprises the issues that arise in the exploitation of the temporal information in digital libraries. We are studying new machine learning techniques to find tendencies, temporal sequences, frequent patterns, etc. from a large document collection. Topic detection and notification of repository changes are two applications we analyse in this research line.

Industrial Applications

• Digital libraries of Newspapers and Journals.
• Knowledge Management in Enterprise Information System.
• Topic detection and Tracking in news streams.