Business Intelligence: Understanding the Basics

Business intelligence (BI) is often called an “umbrella term” and refers to the technologies and activities that extract meaning out of large sets of data. BI activities include querying, reporting, online analytical processing (OLAP) and data mining. Whether you’re sorting columns in spreadsheets or trolling through terabytes of Big Data using sophisticated software, some basic background information is helpful. Here’s an overview:

Efficient data storage: The key to business intelligence

It’s impossible to have a discussion about business intelligence without also discussing database management. The two topics go hand-in-hand because BI queries and the resulting reports are based on data stored in databases. The performance of a database, or its responsiveness to user queries, is directly related to how it stores data (i.e. its structure).

Response time is hugely important to BI applications. Operations such as filtering, join and aggregation benefit from specialized data structures that make it easier to scan large amounts of data stored in relational database management systems (RDBMS). There are four such structures:

  • Indexes – Scanning one column for a particular key value instead of each column in every row makes querying the database faster.
  • Materialized views – Essentially, materialized views are the results of a common query or join operation that are stored as database objects for more efficient access to the data.
  • Partitioning – Partitioning divides tables and indexes into smaller subsets of data so that maintenance operations like loading and backup can be performed on the partition instead of the entire table or index.
  • Column-oriented storage – Instead of organizing data by rows, some RDBMS can organize data by columns. Many non-relational database management systems also use this method of storage to speed up searches.

In addition to specialized data structures, data compression makes for efficient data storage and retrieval. There are five major benefits to compressing data stored in large data warehouses:

  • It shrinks the amount of data that needs to be scanned which reduces the cost of each query,
  • It lessens the amount of storage needed which in turn reduces the cost of storage and backup,
  • The amount of data that can be stored in memory increases because it can be decompressed on demand,
  • Some common queries, like searching for and eliminating duplicate records, can be performed on compressed data, and
  • Compressing data that travels over a network decreases the amount of bandwidth required.

The benefits of data compression enable all types of business intelligence activities, including online analytical processing (OLAP).

Business intelligence queries: Discover the needle in the haystack

OLAP provides users with a multidimensional view of data. The data used by an OLAP application is stored in a multidimensional database that’s fed from existing relational databases.

The foundation of the multidimensional data model is the data cube (“sales”) which is described by dimensions (product model number, date/time of sale, sales region, etc.). Dimensions are described by attributes (product category, model number, year of introduction, etc.) and may have their own hierarchies (for example, cities and states within sales regions).

The OLAP application performs analytical operations such as aggregation, drill-down, filtering and pivoting using dimensions and attributes. In this way, the user can view data from various angles. The biggest advantage of using a multidimensional data model is that data can be processed rapidly.

Data mining is often used in conjunction with OLAP. Data mining is analyzing raw data to uncover patterns that are helpful in answering business questions. Whereas OLAP summarizes data by aggregating data from multiple data sources, data mining takes the analysis a step further and uses algorithms to discover the relationships, the cause and effect, between the data.

Trends in business intelligence: on-demand access for quick decision-making

Self-serve business intelligence applications are all the rage in the business world. Such applications relieve the pressure on IT to keep up with user requests for on-demand access to business critical data. Enabling users to manipulate and analyze data as needed makes sense. These days, users have the skills necessary to use BI tools on their own. Plus, the applications themselves are now much easier for users to navigate than in the past.

Along the same line, mobile business intelligence is a hot topic, too. More and more businesses are taking a harder look at extending their existing desktop BI applications for use on mobile devices as a result of:

  • The success of mobile BI at the executive level of many organizations,
  • The decreasing cost of mobile technology, and
  • The knowledge and experience mobile product developers, BI vendors and IT departments gain through trial and error as the technology matures.[1]

The appeal of mobile BI is hard to ignore. Now that smartphones and tablets can handle the technical requirements of BI applications (both in terms of rendering visual representations of data and bandwidth), experts say that mobile BI is fast becoming a standard in the business world.

Research shows that companies tend to roll out mobile BI applications to senior level executives first and then look at deploying it to the rest of the organization as part of a larger mobile BI strategy. The workers who benefit the most from access to BI on-the-go are salespeople, field service reps, and operations managers – anyone who needs to make business decisions or respond to customers and business partners quickly.

Check out BI Software Insights Resource Guide for IT Pros for more info on how IT can leverage Business Intelligence.

Make Smarter Tech Decisions

Get the latest IT news, trends, and insights - delivered weekly.

Privacy Policy