Time Span and Summarization
How does Time Span differ? Operating data is current and it shows the current status of business transactions. DSS data are a snapshot of the operating data at given points in time. Therefore, DSS data are an historic time series of operating data. We are storing multiple "time slices" of operating data. Inmon (1992) says the DSS data is "time-variant". This characteristic is analogous to putting a time stamp on DSS data when it is loaded in the database or data store.
When does Summarization occur? DSS data can be summarized in the DSS database and it can be summarized by analytical processing software. We may bring some data from a DSS database into a multidimensional data cube to speed-up analysis. Some DSS databases consist exclusively of summarized or what is often called derived data. For example, rather than storing each of 10,000 sales transactions for a given retail store on a given day, a DSS database may contain the total number of units sold and the total sales dollars generated during that given day. DSS data might be collected to monitor total dollar sales for each store or unit sales for each type of product. The purpose of the summaries is to establish and evaluate sales trends or product sales comparisons that will serve decision needs. We may want to ask questions like: What are sales trends for product X? Should we discontinue a product? Has advertising been effective as measured by sales changes? All of these questions can be answered using summarized data. Operational data is not summarized within the database or in any other areas.