Business intelligence applications involve complex queries on very large databases. Users typically view the data as multidimensional data cubes. Computing multidimensional aggregates in large data cubes is a performance bottleneck for many OLAP applications. Calculating the answer of an aggregation query can be too expensive in terms of time and storage space. In this paper we describe some of the problems that can arise in the process of building multi-dimensional applications with Oracle OLAP Option. We pay a special attention to the sparsity of high dimensional data cubes. We present some extensions to the common multidimensional data model witch could solve described problems. They also enable more flexible interface not only for the developer of OLAP application but for the end users too.