BI Beans Metadata Model

The BI Beans metadata model provides an abstraction of OLAP metadata that is stored in Oracle OLAP. OLAP metadata is designed to facilitate sophisticated data analysis.

In the BI Beans model, a measure represents data. For example, a Sales measure might represent the Sales data for an organization. Each measure in an analytic workspace has dimensions, which categorize the data. A Sales measure might be dimensioned by Product, Geography, and Time, for example. This allows for analysis of the Sales of a particular product in a particular location, at a particular time.

Each particular product, time, or location is a member of a dimension. For example, shoes, socks, and handbags might be members of the Product dimension. January, February, and March might be members of the Time dimension. Boston, Tokyo, and Buenos Aires might be members of the Geography dimension.

Of course, when you analyze data, you do not want to see only the data for individual members of a dimension. You often also want to examine aggregations of those members. For example, you want to see the Sales for a quarter, not only for one month. You want to see the Sales for a region, not only for a city. To support this kind of analysis, dimensions have hierarchies, which organize the members into groups.

You might also want to analyze data based on characteristics other than the hierarchical level of dimension members. For example, you might want to see Sales for all of the products of a certain color, without regard to the hierarchical level of the product. Or you might want to see Sales data for the cities that are the responsibility of a certain sales representative. To provide groupings of nonhierarchical data, dimensions have attributes.

The structure of OLAP metadata allows the QueryBuilder and the CalcBuilder to support very sophisticated analytic queries.

Objects that represent OLAP metadata in the BI Beans are defined in the oracle.dss.metadataManager.common package.