<img src="https://certify.alexametrics.com/atrk.gif?account=53pUm1a4KM+2vg" style="display:none" height="1" width="1" alt="">
TRG Blog

Business Intelligence and Analytics - From A to Z (Part 3)

Posted by Rick Yvanovich on

In the previous post, we went throught the six essential Business Intelligence terms (from F to L) you should know. In today' post, we will continue with next six common terms (from M to R).

Business Intelligence and Analytics Terms and Glossary

M - Data Mining

Data mining is the process of identifying patterns and relationships in large data sets. It is one of the major analytics functions of business intelligence. Data mining helps companies forecast future trends based on historical performance patterns.

N - Natural Language Processing

The growing ubiquity of voice control applications like Amazon Alexa, Apple Siri, and Microsoft Cortana has signalled the emergence of natural language processing (NLP), a computer science field focusing on the interaction between computers and human languages.

According to the consulting firm Gartner, 50 per cent of analytical queries will be generated via search or NLP by 2020. A manager then can simply ask: “What was the total sales breakdown by segment last quarter?” in order to get the desired answer.

Natural Language Processing

O - OLAP

OLAP (online analytical processing) is a technology that enables multidimensional analysis of data, i.e. looking at data from different perspectives. For instance, it can display last quarter’s sales of a certain product sold online, and compare it to sales of the same product in the same period of time sold through other channels.   

OLAP forms the foundation of many business intelligence solutions. Its data source is organised in data cubes and stored in data warehouses. Whereas spreadsheet applications like Excel can only display data on a two-dimensional (row-by-column) table, OLAP tools allow users to track data on many dimensions.

Primary operations performed by OLAP are roll-up, drill-down, pivoting, and slicing and dicing. Roll-up is the aggregation of data along a dimension. Drill-down involves going through the details. Pivoting means rotating a data cube in space so users can look at it from different faces. And finally, slicing and dicing involves taking out a certain subset of the data cube and view it from different points of view.

P - Predictive / Prescriptive Analytics

Traditional business intelligence software equipped with descriptive analytics focuses on providing insights into the past performance in order to the question: “What has happened?”

Conversely, predictive analytics from modern BI software aims at forecasting the future trends and answering the question: “What will happen?”

Prescriptive analytics goes one step further and tries to propose courses of action and predict possible outcomes. As such, prescriptive modelling answers the question: “What should be happening?”

Read more: How Grab Uses Data Analytics to Refine New Products

Q- Query and Reporting

In essence, a business intelligence solution is also a query and reporting tool that empowers its users to ask questions (queries) about their data and get the results presented in dashboards or reports.

In the past, (non-technical) business users had to have their reports developed by the IT function, which could take weeks or even months. Modern self-service BI software allows business users to create queries and reports without IT’s involvement.

Download whitepaper "Perfecting Revenue Optimisation For Hotel Chains With Analytics" here

R - Real-time Analytics

Real-time analytics involves being able to provide insights immediately or almost immediately, within a few seconds or minutes after the data arrives. It allows the management to make decisions much more quickly and to react within minutes of market changes.

Some of the enabling technologies for real-time analytics are:

  • In-memory analytics: querying data resided in system memory (RAM) instead of hard disks.
  • In-database processing: data analytics is not a separate application but integrated into data warehouses.
  • Processing in memory: a processor is embedded into the memory. The result is a single chip which can process certain tasks directly within the memory. This architecture helps eliminate the need to move data between the processor and the memory and decrease latency.

This is the 3rd part of our article. Please stay tuned for the 4th part or subscribe to our blog for the latest content about Business Intelligence and Analytics.

Subscribe Now

Topics: Business Intelligence, Analytics

Subscribe to TRG Blog

Upcoming Events

Latest Posts

Most Viewed Posts

Our Editorial Mission

rick yvanovich resized 174

 Rick Yvanovich
 /Founder & CEO/

With TRG International Blogs, it is our mission to be your preferred partner providing solutions that work and we will make sure to guide your business to greatness every day.

Subscribe to TRG Blog

Follow Us