Design process for Big Data Warehouses

Big Data Warehouses differ substantially from traditional data warehouses in that their schema should be based on novel logical models that allow more flexibility than that the relational model does. Furthermore, their design methodology also requires new principles, such as automation and agile techniques, in order to gain both a fast realization and reaction to changes in business requirements. In the paper, we discuss the new features of big data warehouses and present an innovative design methodology based on the key-value model.