SMART: Semantic multidimensional group recommendations

The rising availability of data in the information systems has boosted the challenging problem of queries recommendation, especially in OLAP systems. In this paper, we introduce an innovative oecent a(3)oeoeoe pound system for semantic multidimensional group recommendations to enhance the querying formulation process. Indeed, we describe the problem of group recommendation and define its semantics through introducing our group profiling ontology. Thus, we infer the analysts' ongoing behaviors on our ontological concepts using a weighted summation strategy. Based on our ontological representation, we propose a new method for deriving relevant semantic recommendations (i.e., complete and queries fragments). In addition, an optimization technique for selecting the most interesting visualization of recommendations is proposed. Carried out experiments of our SMART system on real built financial data warehouse highlight encouraging results in terms of precision and recall.