We are actively involved in several research areas related to machine learning, applications to bioinformatics, and algorithms for document image recognition and document engineering. Specific topics of interest include:

  • Declarative languages and representations for learning
  • Learning with relational, logical, and structured representations
  • Kernel methods
  • Artificial neural networks
  • Prediction of protein structure and function
  • Chemoinformatics
  • Document engineering
  • Document image retrieval
  • Graphics recognition
  • E-publishing

The laboratory was founded in 1988 by Prof. Giovanni Soda and over the years has hosted the activities of more than 25 academic staff, researchers, and graduate students.


Artificial Intelligence
Department of Information Engineering (DINFO)
Università degli Studi di Firenze
Via di Santa Marta 3
I-50139 Firenze, Italy
+39 055 479-6363

Machine learning

We are interested in algorithms, architectures, and languages for learning with structured and relational data, using logical representations, kernel methods, neural networks, and probabilistic models.

Applications to bioinformatics

We have been applying machine learning algorithm to the prediction of protein structure and function, and to the prediction of the activity of chemical compounds.

Image Document Recognition

We work on information extraction from both scanned and digital-born documents, with a recent emphasis on the conversion of scientific and technical documents to formats for eBook readers.