Document standards for national and European legislation. Legal document representation, classification and management.
Legal document metadatad standards. IT design of NirEditor and xmLegesEditor, visual drafting environment for legislative documents.
Metadata XML standard for legislative documents.
Techniques of automatic thesaurus mapping for cross-language and cross-collection retrieval.
Legal ontologies for legal reasoning, information retrieval as well as multiligual legislative drafting.

Text categorization, Information Retrieval e Digital Libraries

Machine learning techniques for document classification. Implementation of automatic document classification in the legal domain.
Protocols of metadata harvesting (ex. OAI-PMH) and techniques of information discovery on the Web. Multiligual information retrieval.

Document processing

Document digitalization and processing techniques. Automatic classification and understading of digitalized documents.
Machine learning problems for document logical objects location.
Statistical reasoning in logical objects location. This activity has been carried on within STRETCH, an Esprit project related to digital document processing (project partners: Computer Science Department of the University of Florence and Computer Science Department of the University of Siena, Elsag Bailey, EBI, DCPI, CEA, DFC and DIBE.

Pattern Recognition and Neural Networks

Neural network architectures for pattern recognition: logo recognition, printed characters recognition.
Assessment of autoassociative and recursive techniques.

Knowledge representation

Semantic Networks. Ontologies. Algorithms for semantic networks management. Implementation of such methodologies in the document processing domain.