Welcome to the Knowledge Discovery & Management Group
The
DII research group on Knowledge Discovery and Management focuses on
methodologies and systems for business intelligence (data warehouse,
data mining), business innovation, semantic enterprise interoperability,
collaborative distributed systems. The group has gained years of
experience in data mining field, ranging from pre-processing to model
induction methods. In particular, the group has developed and
continuously improves the Bayes Vector Quantizer (BVQ) algorithm, a
multi-class cost-sensitive learning technique for the design of optimal
classifiers and the definition of optimal sets of features in
risk-based, class-unbalanced domains. The group develops semantic models
for the semantic enrichment of structured objects, like Data Mining
models, processes, and Key Performance Indicators. This research is
aimed at supporting interoperability of managerial activities in
distributed collaborative environments. At present these
interdisciplinary competencies are exploited in the design and
development of two support systems, the KDDVM platform for the
collaborative design and execution of knowledge discovery and e-science
experiments, and the BIVEE platform for Open Innovation in business
ecosystems. Visit the Project and Publications pages for more information on current and past projects and results. Students who like to work on these topics can also visit the Theses page.