Decision Support System (DSS) & Online Analytical Processing (OLAP)

Connelly & Begg (2014) defined business intelligence (BI) as an umbrella term encompassing collection and analysis of data to extract useful trends and information facilitating corporate decision making. Some of the key enablers of BI are data warehouses, online analytical processing (OLAP), and data mining. Data Warehousing is defined as a subject oriented, time-variant, integrated and non-volatile collection of data to support decision making. Data warehousing is subject oriented because it is organized around major subjects of organizations like customers, products and sales. It is integrated because it bring together varying sources of data. It is time-variant as it contains data from several sources with varying timestamps. Finally, it is non-volatile because it is refreshed with the addition of new data instead of replacement.

Online analytical processing (OLAP) is a term that make use of aggregated multidimensional data to provide quick access to information for advanced analysis (Connelly & Begg, 2014). OLAP analysis ranges from basic navigation, slicing, and dicing to complex time series and modeling. Examples of OLAP tools are Microsoft SharePoint, SAP Business Intelligence, and Tableau etc. Performance of OLAP tools is benchmarked using several measures like just-in-time (JIT) information, analytical queries per minute (AQM) etc.

Decision support system (DSS) usually comprise of DBMS running decision making queries on data warehouse (UC Berkeley, 1998). Although DSS and data warehouses are different from online transaction processing (OLTP) systems which supports operations of given organization but one class of DSS queries is sometimes referred as online analytical processing OLAP. For an organization, business intelligence (BI) system will make use of data warehouses and OLAP queries to extract useful trends and information for corporate decision making. A new trend in the industry is to setup a mini-data warehouse, commonly referred as data mart to assist BI needs of a particular department or group of an organization with lots of homogenous data.

With the ever increasing amount of data in data warehouse, it is difficult if not impossible to identify trends and relationship using simple queries and reporting tools (Connelly & Begg, 2014). Data mining is the process of extracting valid, comprehensible, actionable and previously unknown information from large databases to make critical business decisions. Key operations behind data mining techniques are predictive modeling (using supervised learning approach), database segmentation (using unsupervised learning approach to identify segments or clusters), link analysis (relationships, pattern recognition etc.), and deviation detection (identification of outliers using statistical and visualization techniques).

Reference

Connolly, T. & Begg, C. (2014). Database Systems: a practical approach to design, implementation, and management (6th Ed.). Upper Saddle River, NJ: Pearson.

UC Berkeley (1998). Readings in database systems – Data warehousing, decision support & OLAP. Retrieved from http://redbook.cs.berkeley.edu/redbook3/lec28.html