- Business intelligence (BI) is about building a model that answers specific business questions. You start with a question or a set of questions, gather data and data sources, and then build a computer program that uses those resources to provide answers to the business questions you’ve targeted. Business intelligence is about providing the infrastructure for searching your data and building reports.
Business intelligence uses controlled data — predefined, structured data, stored mainly in data-warehousing environments. BI uses online analytical processing (OLAP) techniques to provide some analytical capabilities — enough to construct dashboards you can use to query data, as well as create and view reports. BI is different from data mining and doesn't discovered hidden insights.
- Data mining (DM) is a more generalized data-exploration task: You may not necessarily know exactly what you’re looking for, but you’re open to all discoveries. Data mining can be the first step in the analytics process that allows you to dive into the collected and prepared data to unveil insights that could have gone undiscovered otherwise.
Big data Analytics rely on newer technologies that allow deeper knowledge discovery in large bodies of data of any type (structured, unstructured, or semi-structured); technologies designed for such work include Hadoop, Spark, and MapReduce.
Predictive Analytics emerged to complement, rather than compete with, business intelligence and data mining. You can co-ordinate your pursuit of analytics with business intelligence by using BI to preprocess the data you’re preparing for use in your predictive analytics model. You can point BI systems to different data sources — and use the visualizations that BI produces from raw data — to give you an overview of your data even before you start designing your analytics model. This approach leads to the visualization of large amounts of raw data.