datalab

Concerning data querying, where the aim is to find those objects and data that are better suited to fulfill the information needs of non-expert users, we consider novel techniques taking into account both effectiveness (i.e. "quality" of results) and efficiency aspects of the search process, in order to scale it to large data volumes. Moreover, we also take into account the need to provide the user with simple but powerful tools, able to smooth the processes of query creation/customization and of result interpretation. Until now, we applied our techniques to a plethora of different media types (images, videos, text documents, time series, etc.), to genetic data and to patterns obtained through Data Mining processes (associative rules, clusters, etc.), and recently to genomic data.

Concerning data analysis, our research activity focuses on mining genomic and proteomic data, text documents and unstructured data, on mining data on distributed self-organized architectures (like peer-to-peer systems), on ubiquitous computing (like intelligent sensors or mobile devices networks), and on high-performance multi- and many-core architectures.