Automatic prominence detection in continuous speech, also in a
typological perspective, through the development of models based on
acoustic parameters. Focus and information structure detection in
PATHOLOGICAL LANGUAGE ANALYSIS
Computational analysis of patients' dialogues for an early
identification of Alzheimer's disease.
DISTRIBUTIONAL SEMANTIC MODELS
Geometrical models for computational management of meaning using
Word Space, Semantic Vector Space models or Word embeddings, for Word Sense
Induction and discrimination.
MORPHOLOGICAL ANNOTATION OF LARGE CORPORA AND COMPUTATIONAL TOOLS
FOR MORPHOLOGICAL ANNOTATION
CORIS/CODIS automatic annotation with morphological information to
allow complex queries into the corpus.
Design and development of automatic parsing systems for Categorial
Grammars; development of new models for syntactic analysis -
Unification Categorial Grammars - and design of an Italian grammar,
using also automatic learning processes.
Automatic induction of a set of lexical classes for Italian, to be
used to annotate large corpora; design and development of fast and
reliable Part-of-Speech taggers.
Design and construction of large corpora.
MACHINE LEARNING AND DEEP LEARNING TECHNIQUES
During the last years he has been starting to develop and to apply machine and deep learning techniques to some classical problems in natural language processing, involving, in particular, quantum probability theory.