The business intelligence and analytics architecture divides into three types of analytics.
The traditional analytics that we have been doing for years which covers from raw data from various data sources, clean and summarize the data into data warehouse to prepare standard reports and ad-hoc reports.
These two reports types usually answer the questions of,
- What happened?
- When did it happen?
- Where did it happen?
- Why did it happen?
- How often it happened?
It is often we will drill down to see more detail to answer where exactly is the problem through query drill down or OLAP. OLAP stands for Online Analytical Processing, a technology behind Business Intelligence applications to answer multidimensional analytics queries.
And, finally alerts which gives us indications when we should act and it could be included what kind of actions to be taken. All these are under descriptive analytics.
It includes predictive modelling and analysis, statistical modelling and forecasting. It makes predictions about,
- What may happen next?
- When may it be needed?
- How it may affect?
- What if this trend continues?
- Why is this happening?
It uses many techniques from data mining, statistics, modeling, machine learning and artificial intelligence to analyze historical and current data to determine the patterns and make predictions about future, unknown events and identify risks and opportunities.
It will not tell you what exactly it will happen in the future.
It is dedicated to find what is best action to be taken for a given situation. It is an optimization stage whereby we will look into how to do things better.
These are the main three pillars.
- Descriptive looks into past data and answers “What has happened?”.
- Predictive looks into past and current data and answers “What could happen?”
-Prescription uses optimization and algorithms to advise “What should we do?”.