Development of methods and algorithms of data science, with particular reference to the disciplines of data mining, text mining and natural language processing with machine learning and deep neural network approaches, including unsupervised methods for dimensionality reduction and data clustering in distributed environments, including IoT and data centric sensor networks, and big data.
Concerning text mining and natural language processing, the research interests range from methods of semantic text representation based on efficient non-symbolic and subsymbolic neural language models, algebraic and knowledge graphs, to efficient cross-modal information retrieval architectures and models of texts and images with deep metric learning approaches, and of cross-domain text classification, abstractive and extractive text summarisation.
The research activity in the field of data mining is focused on centralised and distributed data clustering algorithms, in domains of big data and IoT sensors networks, recommendation systems, high-dimensional multi-label classification methods in bioinformatics, with particular reference to genomics and cardiology, and in the development of boosting methods and predictive models in the financial and trading sectors.
