Business Intelligence general refers to identification, extraction or transformation of business data into useful information (reports, charts, graphs etc.) to gain business specific insights like demand forecasts and sales predictions, thus providing better decision making capabilities. It usually refers to the computer based techniques, like reporting, analytics, data mining, benchmarking, predictive analysis etc., but is not limited to them.
As explained by D. J. Power in his work “A Brief History of Decision Support Systems”, there are various tools and technologies that provide Business Intelligence capabilities, and providing an efficient Decision Support System (DSS). His research covers even the basic systems like file drawers, which are used to keep information in organized and readily searchable manner (for small organizations). But in present information age, those kinds of systems seems outdated for requirements of a global organizations, with hundreds of branches across the world, and
generating and processing huge amount of information per hour. In this article, we are focusing only on computer based programs and applications, that consumes and processes the digital information available on organization’s servers, and then generates meaningful results out of it, which provokes better decisions from BDMs (Business Decision Makers), TDMs (Technology Decision Makers) or other IT Pros involved in decision making.
The well-known enterprise analyst organization Gartner predicts a five-fold growth in the Open-Source BI tools product deployment by the end of 2012. They also mentioned in their report on Magic Quadrant for Business Intelligence Platforms 2011, that the growth in BI will be driven by factors like Consumerization of BI and support for extreme data performance with emerging data sources (known as Big Data). And with some recent break-through innovations by the major BI vendors (like SAP’s HANA appliance, Oracle’s Exalytics appliances,
and Microsoft’s BISM model), IT world may expect more surprises coming from the major BI vendors (including but not limited to Microsoft, Oracle, microstrategy, IBM, Information Builders, QlikTech, SAP and SAS).
But irrespective of vendor, all BI solutions have a generic technology stack, with following layers:
· User Interface: This includes the Web based or application based frontend that brings the analysis to the users. It includes the portals (in case of networked or web-based analytics) or the application front end in case of locally deployed BI solution.
· Development and Admin Tools: This comprises of the tools, languages and processes involved in the development and management of BI applications and systems. The difference between BI systems and BI solutions will be covered in another blog. For example, some BI development languages can be MultiDimensional eXpressions (MDX), XML for Analysis (XMLA), Data Mining Extensions (DMX) etc.
· BI Tools: This comprises of the tools (reports, dashboards or otherwise) that enables the users to perform the desired analysis on the underlying data. User access these tools via the User Interface layer discussed above. For instance, Microsoft’s PowerPivot and Power View, SAP’s crystal reports, Jaspersoft, Oracle’s Business Intelligence Foundation Suite are just few examples to name, there are more than 100 of readily usable vendor products available in the market.
· Applications and BI data sources: This comprises of the various sources that keep the information in pre-processed form that can be readily consumed for analysis. This includes models like Online Analytical Processing (OLAP) Cubes or Decision Support Systems, and concepts like Data Mining, Analysis Services, etc.
· Data Integration Tools: This comprises of the various data management tools and concepts like Master Data Management (covering data collection, source identification, schema mapping, normalization, data transformation, rule administration, error detection and correction, data consolidation, data storage, data distribution, data classification, item master creation, data enrichment and data governance) and services like taxonomy services, Data Quality Services,
· Data warehouse platform: This comprises of various data sources, including simple text based files, excel sheets, relational databases, or even complex unstructured data types like audio files, videos, web-logs, click-streams and geo-spatial data etc.